Can nesting behaviour allow reptiles to adapt to climate change?

Abstract

A range of abiotic parameters within a reptile nest influence the viability and attributes (including sex, behaviour and body size) of hatchlings that emerge from that nest. As a result of that sensitivity, a reproducing female can manipulate the phenotypic attributes of her offspring by laying her eggs at times and in places that provide specific conditions. Nesting reptiles shift their behaviour in terms of timing of oviposition, nest location and depth of eggs beneath the soil surface across spatial and temporal gradients. Those maternal manipulations affect mean values and variances of both temperature and soil moisture, and may modify the vulnerability of embryos to threats such as predation and parasitism. By altering thermal and hydric conditions in reptile nests, climate change has the potential to dramatically modify the developmental trajectories and survival rates of embryos, and the phenotypes of hatchlings. Reproducing females buffer such effects by modifying the timing, location and structure of nests in ways that enhance offspring viability. Nonetheless, our understanding of nesting behaviours in response to climate change remains limited in reptiles. Priority topics for future studies include documenting climate-induced changes in the nest environment, the degree to which maternal behavioural shifts can mitigate climate-related deleterious impacts on offspring development, and ecological and evolutionary consequences of maternal nesting responses to climate change.

This article is part of the theme issue ‘The evolutionary ecology of nests: a cross-taxon approach’.

1. Introduction

Ongoing anthropogenic climate changes include increased global surface temperatures, modified regimes of precipitation, extreme weather events (e.g. drought and storm), sea-level rises and ocean acidification. Such changes impose strong impacts on biological systems at all levels from genome to ecosystem [1]. The responses of organisms to climate change thus have attracted great attention from biologists in recent decades [2–8]. Although animals may respond to climate change via pathways such as physiological adjustment and genetic adaptation, behavioural responses may be especially relevant because they can be manifested over short timescales to deal with the threat of climate change on survival and reproduction [9–11]. For example, many species have shifted their distribution upwards or polarwards to track suitable thermal habitats, and ectothermic animals can actively thermoregulate to reduce the elevation of body temperatures that would otherwise result from climate warming [12,13]. Behavioural responses may buffer the impact of climate change and ‘buy time' for species to adapt physiologically and genetically, although this behavioural buffering may sometimes hinder adaptation of physiological traits by weakening selective pressures [9,12,14].

Climate change is affecting reptiles globally, with those from tropical regions being most vulnerable to climate warming because these species often have narrow thermal safety margins and low thermal plasticity [13,15–17]. Oviparous non-avian reptiles are particularly vulnerable to climate change due to (i) high physiological sensitivity to temperature, (ii) a general lack of parental care and (iii) sensitivity of embryonic development and survival to even minor fluctuations in nest environments [18–21]. The embryo is a critical life stage because embryos are highly sensitive to environmental changes that can have far-reaching consequences across subsequent life stages [19,22–24]. Maternal nesting behaviour in oviparous reptiles determines embryonic environments during development, including parameters such as temperature, moisture and oxygen concentration that affect fitness-relevant offspring phenotypes [25–29]. Moreover, maternal nesting behaviour enables females to place their developing offspring in sites that reduce the risks from predators and parasites [30–32]. Consequently, maternal nesting behaviour is expected to undergo strong natural selection, and play an important role in mitigating climate effects, provided that those behaviours are heritable.

Climate change is intensifying the risks faced by developing reptile embryos by modifying exposure to threats such as predation, flooding, overheating, desiccation and pathogens. For example, climate warming may increase soil temperatures to levels that exceed the heat tolerance of embryos, or skew offspring sex ratio in some species with temperature-dependent sex determination (TSD) [33]. Theoretically, nesting females of oviparous reptiles may alter when, where, and how they nest to buffer the impact of climate change [34,35].

To identify or predict the nesting strategies adopted by reptiles in response to climate change, the following kinds of information are important. First, we need to more fully document interspecific differences in nesting phenology and nest attributes before we can predict the evolutionary potential for climate-induced evolutionary responses in reptile nesting behaviour. Interspecific and intraspecific comparisons of nesting behaviour in reptiles may help to identify plausible pathways for evolutionary responses of nesting to different environments and therefore to climate change. For example, phylogenetic conservatism may render some aspects of nesting behaviour invariant within lineages, enabling us to infer that deviations from that widespread pattern of behaviour are unlikely to be targets of rapid evolutionary change [36]. Alternatively, convergent or parallel evolution may generate similarities between sympatric distantly related species in nesting biology [37].

Second, identifying the degree and triggers of plasticity (e.g. spatio-temporal variation) in traits associated with nesting may help us to predict ecological responses of nesting behaviour to climate change. Plasticity of nesting reflects the ability to modify nest design in response to environmental stimuli, including spatial and temporal variations in biotic and abiotic factors such as predation risk, temperature and precipitation [38,39]. Moreover, intraspecific variation in nesting behaviour can shed light on microevolutionary potential in response to climate change, as phenotypes of different populations may have evolved to match local environmental optima [40,41]. For example, Australian water dragons (Intellagama lesueurii) nested in more open areas and at shallower depths in cooler latitudes and elevations [35,42].

Third, longitudinal observations are needed because such studies provide direct evidence on how species have changed their nesting behaviour in response to climate change. For example, 10 years of nest monitoring showed that montane scincid lizards (eastern three-lined skink, Bassiana duperreyi) now nest earlier and deeper in response to increasing ambient temperatures [40].

Fourth, laboratory experiments that manipulate weather variables may provide cues as to how a species will respond behaviourally to climate changes. For example, female three-lined skinks did not change nesting behaviour when they experienced warming conditions during the period of uterine retention of eggs, but they dug shallower nests in response to warmer conditions during the oviposition period [43]. Therefore, exploring interspecific and intraspecific differences in nesting behaviour of reptiles and their consequences for embryonic survival and offspring phenotypes can clarify how nesting behaviour responds to climate change. These responses may, in turn, affect nest environments of reptiles and therefore both developmental success (survival to hatching) and offspring traits (e.g. time to hatching, offspring morphology or behaviour or sex) that have long-term effects on population dynamics.

The topic of how nesting behaviour responds to climate change has received increasing scientific attention. Mainwaring et al. [32] summarized ecological threats facing nest-building vertebrates and their potential adaptive short- and long-term responses under climate change, while other reviews have explored issues associated with nesting behaviour and its response to climate change in specific lineages or species of reptiles [44–48]. In this review, we focus on the nesting behaviour of non-avian reptiles and how reptiles may adjust nesting strategies in response to climate change. First, we provide an overview of nesting behaviour and nest design of reptiles, and summarize evolutionary adaptation and plasticity of nesting in response to environmental variation in reptiles by identifying interspecific and spatio-temporal variations in nesting behaviour. Second, we discuss the nature of challenges that climate change imposes on reptile nests and embryos. Third, we consider changes in nesting strategies in response to rising temperatures and how such changes may buffer the impacts of climate change. Finally, we identify knowledge gaps and suggest future research priorities.

2. Nesting behaviour and nest design of reptiles

Non-avian reptiles consist of five distinct groups that differ significantly in ecological traits (turtles, tuatara, lizards, snakes and crocodilians). Oviparous reptiles build two main nest types: mound nests and hole nests [29,49,50]. Mound nests are constructed above ground with vegetation, leaf litter, woody debris, sand or soil. Such nests are relatively rare in lizards and snakes (but see the diamond python Morelia s. spilota; [51]), but characterize almost 80% of crocodilians, including all eight alligators, at least 12 crocodiles and the false gharial (Tomistoma schlegelii) [52,53]. The size of nest mounds is determined by the amount of material accumulated by the female, depending on species, sites, and availability of nesting materials [51,54].

Hole nests are generally shallow scrapes, underground burrows or cavities with an expanded chamber containing eggs at the end dug by females, or existing burrows or crevices selected by females. Hole nests have been observed in the tuatara (Sphenodon guntheri) [55], snakes [56,57], some crocodilians [58,59], and many lizards [60–62] and turtles [63–65]. The structure of hole nests is similar in tuatara, lizards and snakes, consisting of a tunnel leading to the egg chambers [55,56,62]. By contrast, hole nests of many turtles are flask-shaped without a tunnel, which may be due to their specialized body shapes and their method of digging holes with their hind legs [65,66]. The structure of hole nests in crocodilians is similar to that in turtles, because crocodilians construct their hole nests with the hind legs [44,52]. In hole nests, nest depth is highly correlated with nest temperature and moisture. As the depth increases, the temperature range in the nest usually decreases while the soil water content increases [67,68], although the latter gradient is reversed after rainfall [39].

Additionally, some lizards and snakes place their eggs in existing holes or crevices above the ground, or under cover (e.g. litter, logs, rocks) at the ground surface. For example, most oviparous geckos and some snakes lay their eggs in crevices of rocks or anthropogenic structures, tree holes and debris [69,70]. Some lizards and snakes even deposit their eggs inside nests of other animals (e.g. other reptiles, ants and termites). For example, little brown skinks (Scincella lateralis) oviposit in nests of American alligators (Alligator mississippiensis) [71], an African lizard (Tetradactylus africanus africanus) lays its eggs in nest mounds of the ant Anochetus faurei [72], and other reptiles nest within the nesting burrows of monitor lizards (Varanus panoptes, Varanus gouldii) [73].

Our collation of published data on depth of underground nests across reptile taxa shows that median values of nest depths are strikingly similar (approx. 15–30 cm) despite the substantial variation in body sizes among reptile taxa. The major exceptions involve extremely deep nests in monitor lizards (100–300 cm) and sea turtles (approx. 60–80 cm) (figure 1a). Nest depth is positively correlated with body size in turtles because turtles use their rear limbs to dig nests, and larger turtles have longer legs (nest depth = 0.602*body size +1.453, F_1,38 = 245.12, p < 0.001, R² = 0.866; figure 1b). As a result of diverse nest designs in lizards, nest depths vary significantly among species, and the relationship between nest depth and lizard body size is not as tight as seen in turtles (nest depth = 1.060*body size + 17.616, F_1,30 = 3.754, p = 0.06, R² = 0.111; figure 1b). However, nest depth is not associated with female size in species constructing burrow nests. For example, tuatara and some tortoises construct long burrows but dig these into sloping ground with the result that the final nest is shallow despite the long burrow [74,75]. Nest depth was not significantly correlated with female body size in burrow nests of the desert-dwelling toad-headed agama (Phrynocephalus przewalskii) [62].

Figure 1.

Nest depths in reptiles, based on a collation of published data. (a) Nest depths of different reptile taxa (tuatara, n = 1; snakes, n = 9; lizards, n = 32; turtles, n = 40; crocodilians, n = 6). Data were collected from the literature. A mean or median value of nest depth was used in the analysis when multiple data sources or ranges of data were available. The top and bottom of the boxes indicate the 75th and 25th percentiles, respectively; the line inside the box indicates the median; whiskers (error bars) above and below the box indicate the 90th and 10th percentiles, respectively; and black circles indicate outliers. (b) The relationship between nest depth and body size in turtles and lizards. Nest depth is positively correlated with body size in turtles and lizards. (Online version in colour.)

Divergent nest-site choices and nest structures lead to different nest environments among different lineages of reptiles, based on our collation of published data on nest temperatures. The mean nest temperature was 16.9°C for tuatara, 25.3°C for lizards, 24.6°C for snakes, 28.1°C for turtles and 30.8°C for crocodilians (table 1). Although the number of species with recorded nest temperatures is low, these mean nest temperatures correspond well with the thermal ranges suitable for successful embryonic development in different groups of reptiles, with a narrow range of low temperatures in tuatara, a narrow range of high temperatures in crocodilians, and a wide range of temperatures in lizards, snakes and turtles [23]. In addition, these nest temperatures correlate with the embryo thermal tolerances of these groups (i.e. tuatara have the lowest heat tolerance and crocodilians have the highest heat tolerance, with lizards, snakes and turtles in between) [76]. The range of thermal variation among nests was lower in crocodilians than in other lineages, especially in lizards (table 1).

Table 1.

Nest temperatures in different lineages of reptiles. Data were collated from published literature, N, the number of species. We averaged the nest temperature if there were multiple reports of nest temperatures within the same species. (Temperatures shown in °C.)

	mean temperature	s.e.	temperature range	N
tuatara	16.9	—	5.6–28.7	1
lizard	25.3	1.18	9.5–44.5	10
snake	24.6	2.00	16.0–34.5	6
turtle	28.1	1.00	20.3–36.5	7
crocodilian	30.8	0.35	28.0–35.1	5

(a) . Spatial variation of nesting behaviour

Female reptiles tend to construct relatively cool nests that avoid lethally high temperatures in warm areas, but relatively warm nests that hasten embryonic development in cold areas [77,78]. To achieve this matching between nest structure and ambient weather, females from regions with different thermal environments have different nesting behaviours in terms of nest timing, nest-site selection and nest structure [35].

First, female reptiles from warm climates usually nest early in the summer season, thereby avoiding extreme high temperatures. For example, females in warm-climate populations of the American snapping turtle (Chelydra serpentina), yellow mud turtles (Kinosternon flavescens), and Australian water dragons nested earlier than did their cooler-climate conspecifics [35,79]. A common garden experiment in outdoor ponds observed a similar pattern in five populations of painted turtles (Chrysemys picta) [80].

Second, female reptiles from a warm climate may choose more shaded nests than their counterparts from a cold climate. For instance, painted turtles in a southern population reduced nest temperatures by nesting close to water [81]. The American snapping turtle nested in shaded areas at low latitudes, but in open areas at high latitudes [82], as did the velvet gecko (Amalosia lesueurii) [70]. Analogously, Australian water dragons chose nest sites with more shade cover at lower latitudes than at higher latitudes [35]. In addition, females of an invasive lizard (Anolis cristatellus) in urban sites experienced increased temperatures compared with their counterparts in the forest; and therefore urban-dwelling females chose nest sites closer to trees that provided shade [83].

Third, female reptiles in warm climates may build deeper nests than do their counterparts in cold climates. For instance, female painted turtles from low latitudes dug deeper nests than those from high latitudes [80]. Analogously, the low-elevation water dragon constructed deeper nests than did their high-elevation counterparts [42]. Australian water dragons dug deeper nests in urban areas than in natural riparian habitats [84].

Nonetheless, some investigations have not found significant spatial variation in nesting behaviour of reptiles. For example, nest design of the American alligator did not differ between northern and southern populations [85]. Moreover, hydric conditions may also affect spatial variation in nesting behaviour. For example, female iguanas from an island with high precipitation began nesting five to six weeks earlier than those from an island 140 km away with low precipitation [86]. Loggerhead sea turtles (Caretta caretta) from the Florida Atlantic coast predominantly nested mid-beach in hot and humid subtropical sites, but further from the water's edge in warm and dry temperate-climate sites [87].

Nest-site selection by females may also differ at small spatial scales, affecting nest environments and thereby developmental success of embryos and phenotypic traits of offspring. For example, green turtles (Chelonia mydas) and loggerhead turtles nested at two beaches with different types of sand (dark versus light sand) that generated significantly different nest temperatures and therefore different rates of embryo survival (i.e. hatching success) [88,89]. Similarly, females laid their eggs in two types of nests with different thermal environments in grass snakes (Natrix natrix) from cool northern Europe (compost piles versus manure heaps), and water pythons (Liasis fuscus) from the Australian tropics (burrows of varanid lizard versus cavities beneath tree-roots) [90,91]. In addition, sympatric species may select different nest sites and therefore enable their progeny to incubate in divergent nest environments. For example, some Australian montane skinks (Nannoscincus maccoyi, Bassiana duperreyi) nest in different habitats and show interspecific divergence in nest temperatures allied to interspecific divergences in thermal tolerances of embryos [92].

(b) . Temporal variation of nesting behaviour

In reptiles, temperature is a primary factor that drives the onset of breeding, which often is correlated with winter or spring temperatures [79,93]. Higher-than-usual temperatures in the winter and spring may permit earlier emergence from hibernation and accelerated vitellogenesis and ovulation that may drive earlier nesting [94]. For example, earlier nesting has been found after warm winters in painted turtles [95], or in warmer springs in painted turtles, snapping turtles and tuatara [74,95,96].

Most oviparous reptiles lay their eggs in spring and summer, a period over which ambient temperature can fluctuate substantially [97]. However, many female reptiles do not adjust nest-site selection in response to seasonal variation in temperatures. For example, female brown anoles (Anolis sagrei) nested at similar relatively cool locations across the season, with nest temperatures driven by seasonal variation in ambient temperatures [38]. Australian water dragons constructed nest sites with similar canopy openness and nest depth despite gradually increasing air temperatures during the nesting season [98]. Female jacky dragons (Amphibolurus muricatus) consistently chose sites with lower-than-average canopy cover, resulting in relatively warm nests [99]. Tuatara did not vary their nest depth across the year, and slightly deeper nests later in the breeding season is likely due to easier digging in reused tunnels by later-nesting females rather than thermal cues [100].

Some reptiles show nest fidelity across years, suggesting little plasticity in maternal nest-site selection. For instance, about 25% of female tuatara chose locations containing conspecific cues from previous nesting seasons as nests, and 93% of female tuatara nested in the same rookery at least twice in 5 years [100]. Similarly, females used the same communal nests annually in the Australian velvet gecko (Amalosia lesueurii) and eastern three-lined skink, probably based on the existence of eggshells from previous seasons [101,102]. Nest-site choice was consistent within nesting seasons or among several years in the hawksbill sea turtle (Eretmochelys imbricata), the green turtle and the painted turtle [103–105]. However, despite nest-site fidelity, females of the green turtle adjusted nesting behaviour through laying later nests closer to each other than earlier nests [104].

In addition to temperature, rainfall and soil moisture can also affect nesting behaviour in reptiles. For example, rainfall triggered the onset of the nesting season in some turtles experiencing drought [106,107], and the timing of nesting was associated with the magnitude of rainfall during the previous wet season in the pig-nosed turtle (Carettochelys insculpta) [108]. In addition, females nested deeper in drier years in the deep-nesting monitor lizard (Varanus panoptes) [39].

3. Climate change challenges to reptile nests and embryos

Climate change directly increases temperatures and alters hydric conditions in reptile nests. Those abiotic changes may dramatically modify the fate of embryos in the nests unless females can adjust their nesting locations and nest structures [43]. Nest temperatures of the eastern three-lined skink continuously increased by ca 1.5°C from 1997 to 2006, despite the adaptive adjustment of nesting behaviour [109]. Nest temperatures became warmer and more variable in artificial habitat from the year 2001 to 2009, but not in natural habitat in the long-tailed skink (Eutropis longicaudata) [110]. For geckos that expose their eggs directly to air, nest temperatures are closely related to air temperature and are expected to increase under climate warming [101]. In addition, the urban heat island effect has caused higher mean and extreme temperatures in lizard nests from city suburban areas than in adjacent natural habitats (e.g. mean nest temperature: suburban versus forest: 28.4 versus 26.8°C for Anolis species) [111,112].

Climate change also can affect the nest environments of turtles and crocodilians that nest beside water bodies. El Niño events reduce average precipitation and increase mean temperature, which may in turn influence temperature and moisture levels of nests in South American turtles [113]. Mean temperatures in many sea turtle nests are typically around 29–32°C, within the optimal developmental range, and increases of 2–3°C in air temperatures (as is predicted by climate-change models) would decrease nest emergence success and hatchling locomotor performance in current nest sites [114]. The nests of American alligators will warm by 1.6–3.2°C by the end of this century if maternal nest-site choice behaviour does not adjust to changing ambient conditions [85]. Increased ocean temperatures are expected to produce an 18–60 cm rise in sea level by 2100, and more extreme weather events such as hurricanes and typhoons [1], which may cause significant loss/erosion of nesting beaches used by sea turtles [115–117]. The impacts of sea-level rise and extreme weather on nesting sites of sea turtles are species-specific. Species with lower nest-site fidelity (e.g. leatherback turtles, Dermochelys coriacea) would be less vulnerable to storms than those with higher site fidelity (e.g. hawksbill turtles) [116]. The sea-level rise driven by global warming is also a major threat to the area of suitable nesting habitat for mound-nest-building crocodilians such as saltwater crocodiles, Crocodylus porosus [45].

Extensive work has revealed that incubation conditions can dramatically affect incubation duration, hatching success, hatchling phenotypes and survival in many species of reptiles [19,24,118]. The generality of that response suggests that climate change will have negative consequences on embryonic development and offspring phenotypes in reptiles, such as increased egg mortality, decreased body size, biased sex ratio and so on.

(a) . Developmental rate

Increasing nest temperatures can profoundly affect the developmental rate of reptile embryos. As mean ambient temperature during the nesting season increases, incubation duration decreases significantly, and hatching date has shifted earlier in the loggerhead sea turtle and the American crocodile (Crocodylus acutus) over the past several decades [119,120]. In addition, climate change also increases thermal variability in addition to mean temperature, and increased thermal variability can shorten incubation duration compared with constant temperatures [121,122]. Briefer incubation may benefit offspring. For example, the shortened incubation duration may extend the activity period of hatchlings before winter and thus increase the growth and survival of earlier hatchlings [123,124]. However, faster developmental rates also might induce detrimental effects on hatchlings. Numerous studies on reptiles have demonstrated that faster development at warmer temperatures often results in smaller body size at hatching [24,125]. The small hatchlings resulting from higher-temperature incubation may be less able to compete for food or to evade predation, although such fitness-reducing effects [126] are not universal [127]. Additionally, flooding or rainfall can affect embryonic development rate directly or indirectly by interacting with nest temperatures [128]. For example, eggs of painted turtles had longer incubation duration in watered nests without any change in nest temperatures because of water saturation-induced hypoxic conditions that disrupted embryonic development in eggs [129].

(b) . Egg survival

Climate change may also impose significant challenges on egg survival, which could directly influence population viability in reptiles [130]. Increased thermal variability due to climate warming may cause increased mortality of developing embryos in freshwater and marine turtles and potentially also in other species that construct shallow nests, due to periods of lethally high nest temperatures [131–133]. Human activity may worsen this process; egg survival of the long-tailed skink decreased in nests from warm artificial habitats but not in nests from natural habitats as the air temperature increased [110]. As well as direct impacts of increasing temperature, sea-level rise and beach flooding caused by climate warming also have a negative impact on egg survival in reptiles nesting near water bodies. For example, increasing nest water content decreased hatching success and emergence success in loggerhead sea turtles [134]. Sea-level rising can flood green sea turtle nests and kill the developing embryos [135]. Nonetheless, the influence of future climate change on hatching success of sea turtles is predicted to differ among regions, with negative impacts on tropical sea turtle populations due to severe reductions in hatchling production, but neutral or even positive impacts on temperate populations [136]. Climate-change models also predict increasing frequencies of extreme events that can result in extensive flooding or fires, which potentially could be fatal to developing embryos inside nests [127].

(c) . Offspring phenotypes

Models of climate warming predict a severe impairment of hatchling quality (e.g. body size and behavioural performance) and a highly biased sex ratio in TSD reptile species [89,137,138], but longitudinal studies that can provide direct evidence of phenotypic changes induced by climate change are still rare. Nonetheless, numerous studies have documented the effects of incubation temperatures on offspring phenotypes in reptiles ([24,114]; see reviews in [118,128,139]). High developmental temperatures at the embryonic stage produce small hatchlings that might have poor locomotor and growth performance [125,140–142]. High developmental temperatures also can impair behavioural (e.g. learning ability) and physiological (e.g. immune capability) traits of hatchling reptiles [143–145]. Such negative effects of high developmental temperatures on the behaviour and physiology of hatchlings will eventually affect offspring fitness. In addition, such effects might accumulate and persist for multiple generations [146,147]. Therefore, climate change likely imposes detrimental effects on individual fitness and therefore population growth and viability in some species.

Among the phenotypes of offspring, the most sensitive parameter may be offspring sex ratio in reptile species with TSD because even slight temperature increases (1–2°C) can induce a sex-ratio bias that may be sufficient to trigger eventual population collapse [148]. Extremely female-biased sex ratios have been observed in wild populations of green turtles from the Great Barrier Reef [149]. Theoretically, reptiles with a TSD pattern of Ib and II (i.e. tuatara, some lizards and crocodilians), which produce males at warmer temperature [150], might be even more vulnerable to global warming because climate warming may induce male-biased populations that can severely reduce population growth [151]. For example, sex ratios of nests in tuatara were male-biased (64% males) in the relatively warm breeding season of 1998/1999 [151], and models generated in that study predicted that 100% of tuatara hatchlings would be male by the mid-2080s under extreme regional climate change [148]. By contrast, it has been projected that temperature rises of 1.6–3.2°C in 2090–2100 may skew the sex ratio of American alligators to 98% females [85]. In addition, the increasing temperature variability under climate change can intensify TSD effects in species with shallow nests [131,152]. Remarkably, climate change may also modify sex determination in some species with genotype-dependent sex determination (GSD), because high or low incubation temperatures can override the genetic mechanism in some GSD species, including the Australian central bearded dragon (Pogona vitticeps) [153] and the eastern three-lined skink [154]. For example, sex reversals (high incubation temperatures cause genetically male animals to develop into females) have been found in hot nests in natural populations of P. vitticeps [155], and likely occur in other lizards with a similar sex-determining system [156].

In addition to these direct impacts of climate change on nests and embryos, climate change may affect the environment of reptile nests indirectly. Climate change affects rainfall patterns and thereby the magnitude of predation risk on eggs, given that rainfall during or after nesting may reduce nest predation in turtles [106], and increased rainfall increased rates of ant predation on the eggs of the lizard Anolis limifrons [157]. Climate change also can lead to macroalgae blooms (Sargassum spp.) that might invade nesting areas of sea turtles in tropical and subtropical regions. Sargassum cover may not only affect the access of female sea turtles to their nesting beaches, but also alter nest temperatures and induce hypoxic conditions, thereby affecting embryonic survival or offspring sex ratios [158–160]. Changes in temperature and moisture can have dramatic effects on rates of spread of invasive weeds, potentially shading (and thus cooling) natural nest sites for many reptiles. The invasive common reed (Phragmites) shaded out turtle nests after laying in Ontario, increasing incubation period by depressing nest temperatures [161]. By contrast, anthropogenic habitat under climate change also may provide new nesting sites for reptiles. For example, American crocodiles expanded their nest area to the canals and ditches that were dug for drainage, navigation or cooling purposes in Florida, which led to increased survival of hatchlings [44,162].

Other indirect effects of climate change on developing reptilian embryos may involve flow-on impacts of changes to maternal thermoregulation or energy budgets. For example, the phenotypes of hatchling reptiles may be affected by maternal body temperatures prior to oviposition [144], which in turn may be affected by climate change [12]. The total number of eggs in a nest also can affect hydric transfer among those eggs and thus, the phenotypes of hatchlings [145]. If climate change modifies food availability for a reproducing female, the consequent shift in egg size or number might change the developmental environment for embryos via shifts in clutch volume [146]. By modifying parameters such as length of the mating season and thus opportunities for multiple-mating and consequent sperm selection by females, climate change can also affect genetic traits of developing embryos [147] and hence, potentially, their responses to abiotic conditions inside the nest.

4. Nesting strategies in response to climate change

Although climate change can impose significant impacts on embryonic development and hatchling phenotypes, diverse temporal and spatial variation in the nesting behaviour of reptiles implies that females are capable of adjusting their nesting behaviour to mitigate the impact of climate changes. These adjustments, which may involve nest timing, nest-site selection and nest structure, can potentially buffer the negative impact of climate change [26].

(a) . Adjusting the timing of nesting

Female reptiles nest earlier in warmer regions and years, suggesting that females may be able to nest earlier in the season in response to climate warming. Consistent with that scenario, climate warming appears to have advanced nesting date in some species but not others [163]. For example, a long-term field study showed that five species of freshwater turtles from the northern USA and Canada have nested increasingly earlier over the past 36 years, with the northernmost population of the red-eared slider turtle (Trachemys scripta) exhibiting the most advancement in the onset of nesting [73]. Over the period 1997–2006, the eastern three-lined skink from Australia nested four weeks earlier in response to 1.6°C warming [109]. The American crocodile also has advanced the timing of nesting in response to rising temperature [120], as did the Chinese alligator (Alligator sinensis) [164]. This earlier nesting represents an adaptive response that might buffer the detrimental impact of climate change on embryos [46]. Earlier nesting in a warmer spring resulted in even or female-biased offspring sex ratios in tuatara, mitigating the male-biased effect of rising temperatures [74], although it has also been suggested that nesting later may compensate for the male-biasing effects of warmer air temperatures [148].

By contrast, earlier nesting may only partly compensate for climate change in other species. For example, earlier nesting by the eastern three-lined skink could not fully buffer the observed 1.6°C warming [109]. In painted turtles, earlier nesting in warmer years did not prevent an alteration in sex ratio due to increased summer temperatures [95]. Part of the reason for such effects may be that climate change can alter the seasonal progression of soil temperatures, such that a given temperature regime at the time of oviposition no longer predicts subsequent temperatures during incubation in the same way as before; and phenotypes of hatchling reptiles may be affected by such seasonal changes in temperature during incubation [150]. Moreover, earlier nesting and a prolonged reproductive season due to climate warming may enable females to lay more clutches in a given reproductive season, which can facilitate population growth [95,165]. However, under some circumstances the extra clutch produced later in the season, owing to earlier nesting, could be too late for eggs to complete development. Such futile energy allocation suggests a possible negative effect of earlier nesting [46].

(b) . Changing nest sites

Female reptiles are selective in where they lay eggs, tending to choose nest sites that maximize hatching success and offspring fitness [26,35]. It is thus a reasonable expectation that nesting females are able to select specific nest sites with suitable environments for embryonic development to buffer the effects of climate change [32,48,166,167]. Females choose suitable nest sites based on parameters such as temperature, vegetation cover, canopy openness and moisture content [13,168–171]. Female reptiles may select open nest sites in cooler regions but shaded nest sites in warmer regions, thereby compensating at least partly for temperature variations across latitudinal and elevational clines [35,42,78]. Females may well be able to do the same thing in the face of climate change. Reptile species with TSD can buffer the negative effects of warming climate on offspring sex ratio if they select shaded nest sites [35,148,172]. For example, green turtles in Costa Rica selected cooler nest sites under trees rather than in grassy areas in response to recent climate warming, producing both sexes of hatchlings even during the extremely hot season of 2015–2016 [104]. Similarly, Florida softshell turtles (Apalone ferox) nested in shaded sites that depressed mean nest temperatures by 2°C, potentially buffering the impact of a warming climate [48]. By contrast, species or populations that already nest in shaded areas [35,67] may be vulnerable to climate warming because they already use the coolest microclimates [48]. Accordingly, maintaining habitat diversity in nesting areas may be a useful management tactic to provide opportunities for females to choose suitable nest sites to mitigate the impact of increasing global temperature [48].

Remarkably, in response to the ongoing global warming, Kemp's ridley sea turtles (Lepidochelys kempii) have shifted their nesting range from tropical regions to further north (away from the equator) in the past 30 years, thereby achieving cooler nests [126]. Sea turtles may be able to nest successfully by colonizing new beaches even if some existing nesting beaches disappear or become thermally unsuitable, a flexibility that may be critically important under climate change [115]. American crocodiles in Florida show a bet-hedging strategy of nesting behaviour. Females nest in sand mound nests that resist flooding but are vulnerable to desiccation, or in hole nests on marl creek banks that resist desiccation but are prone to flooding. This diversity may guarantee successful hatching of some nests even under extreme conditions induced by climate change [44].

Nonetheless, some features of nesting behaviour in reptiles may hinder nest-site selection in response to climate change. First, female reptiles that select their nest sites based on specific cues such as vegetation coverage (e.g. ‘nest in an open area') may stick to these cues despite seasonal thermal variation [28]. Warming experiments demonstrated that gravid plateau fence lizards (Sceloporus tristichus) did not modify their nesting behaviour under rising temperatures [43]. Second, nesting fidelity has been reported in tuatara, some turtles, lizards, snakes and crocodilians [100,173]. This feature of nesting behaviour may be adaptive today, but could be detrimental in the future if females do not relocate to new nest sites when climate change makes their original nest sites unsuitable. Third, communal nesting is widespread in reptiles [174] and may be associated with little plasticity in maternal nest-site selection because females use social rather than abiotic cues when selecting nest-sites. Such criteria might reduce the ability to buffer climate change impact through altering nesting behaviour [101].

(c) . Modifying nest structure

In addition to nest timing and location, female reptiles may be able to adjust nest structure to buffer the impact of climate change. Female reptiles can modify nest depth and thereby the mean and variation of temperatures within the nest ([68,175,176], but see [177]). Mothers may construct deeper nests in warmer climates, thereby avoiding extreme temperature fluctuations that are common in shallow nests [46,176]. For example, the eastern three-lined skink not only nested earlier, but also dug 15-mm deeper nests as the local climate warmed over the period 1997–2006 [109]. In addition, female reptiles may nest deeper in response to soil moisture rather than temperature. For example, female deep-nesting monitor lizards dug deeper nests in drier nesting seasons [39]. Diamondback terrapins (Malaclemys terrapin) improved hatching success by nesting deeper during an unusually hot and dry breeding season [34].

Our understanding of how nest depth affects embryonic development and offspring fitness is limited and the conclusions are mixed. For example, deeper nests may mitigate the negative effect of increasing temperatures on embryonic development and offspring phenotypes [34], but may not fully compensate for climate warming [109]. Manipulative experiments found that nest depth within a biologically relevant range had no effect on incubation regimes and offspring traits (e.g. size, sex ratio and survival) in the painted turtle [177]. It is noteworthy that nest depth is constrained by female body size in many reptile species (most turtles and lizards) that use rear limbs to dig their nests (figure 1), or the ability of hatchlings to escape from a deeper nest. These may restrict the opportunity for these species to construct nests with different depths in response to climate change [136,148,178].

5. Future perspectives

Although female reptiles may be able to change their nesting behaviour to buffer the impact of climate change, our understanding of how nesting behaviour responds to climate change and its ecological and evolutionary consequences is limited. One reason for that problem is the limited scope of studies on nesting behaviour in reptiles. For example, most of our current knowledge comes from studies on turtles and crocodilians in which nesting behaviour can be easily observed. We know far less about nesting behaviour in squamates (lizards and snakes), the most speciose and diverse lineage (containing 95% of total reptile species; http://www.reptile-database.org/ ; figure 1). The ecological driving forces and potential for rapid evolutionary change in maternal behaviour also remain unclear. Here, we outline several research areas that might help to clarify the role of nesting behaviour in buffering impacts of climate change.

(a) . Changes in the nest environment under climate change

Monitoring changes of nest environments through time is a first step to evaluate potential nesting strategies in response to climate change. For studies at global and regional scales, we can download environmental data from online climate databases [179]. However, environmental parameters collected by this method may represent soil environments at some specific depths (e.g. 5 cm underground) rather than exact nest environments. For studies on individual species, we can use data loggers to record environmental parameters of field nests at different intervals from minutes to days [135]. However, this method is labour-intensive and depends on the feasibility of locating nests in the field. As we mentioned above, nest environments have been relatively well documented in tuatara, turtles and crocodilians, but not in most lizards and snakes. One major logistical challenge is to locate nests in lizards and snakes with cryptic nesting behaviour. Once we find a nest, it is easy to record nest temperatures using miniature dataloggers (e.g. i-Button temperature loggers), but more difficult to monitor other environmental factors such as water and oxygen contents. Another major challenge is to monitor nest environments over a period that spans many years, which is essential for documenting the impacts of climate change on nest environments. Several impressive examples of such studies include long-term records of nest environments in the painted turtle from North America [180] and the eastern three-lined skink from Australia [109]. Collecting such data is time-consuming, and needs careful experimental design and organization with continuous financial support. Those constraints make such cases very rare. Nonetheless, monitoring environmental parameters of real reptile nests (even over a short period) is extremely valuable for studying nesting ecology. The resulting information is critical not only for designing ecologically relevant experiments to understand how and why reptiles construct specific nests, but also for comparative studies in the future to identify changes and consequences of nest environments.

(b) . Nesting strategies used by reptiles to buffer climate change

Although we know that female reptiles can mitigate the effects of climate change by advancing nesting phenology, and constructing deeper nests in shade [48,79,80], empirical evidence remains scanty, especially in lizards and snakes [109]. Therefore, future studies should pay more attention to the following questions: (i) what nesting strategies are used by lizards and snakes to buffer abiotic variations and especially, climate change effects?; (ii) how does nesting behaviour differ among and within species in response to climate change, and to what degree is that variation driven by heritable traits versus behavioural plasticity?; (iii) do female reptiles modify other nest traits (e.g. burrow shape of hole nests, and nesting materials of mound nests) in addition to nest depth in response to climate warming?; (iv) which environmental cues are used by females to choose nest sites that may buffer the negative effect of climate change?; and (v) how quickly can a reptile individual or population shift their nesting strategies adaptively under climate change? To answer these questions, field observations and manipulative experiments are essential to disentangle the factors determining where females lay their eggs, and predicting how nesting behaviour will be affected by climate change.

(c) . Ecological and evolutionary consequences of nesting strategies

The ecological and evolutionary consequences of nesting strategies in response to environmental or climate change are largely understudied in reptiles. Some priority topics are listed below.

First, although it is known that female reptiles can adjust nest timing, location and structure in response to climate change, the ecological consequences of these changes in nesting behaviours have yet to be explicitly revealed. Some studies have demonstrated that these behavioural adjustments may induce significant changes in offspring phenotypes like body size, locomotor performance and sex ratios in species with TSD [85]. However, whether these phenotypic changes have long-term effects on offspring fitness and therefore population dynamics is unknown. To answer this question, we need long-term studies to quantify the effect of changes in nesting behaviour on the survival and reproduction of offspring. This kind of study may be impossible in long-lived species (such as most turtles and crocodilians), but would be feasible with lizards and snakes with short lifespans. Recently, warming experiments on lizards in semi-natural conditions have been conducted to reveal life-history responses to climate warming [181,182]. Such studies can be easily extended to the responses of nesting behaviour to climate warming and the consequent effects on offspring survival and reproduction. In addition, changes in nesting phenology may disrupt species interactions if the prey, parasites and predators of a reptile species respond differently to climate change. Such trophic and host–pathogen dynamics would further complicate interpretations about the adaptive significance of changes in nesting behaviour under climate change. Studies on such relationships should be encouraged.

Second, nesting behaviour is selected not only for successful embryonic development, but also for maternal and offspring fitness [26,29]. Selection pressures differ among life-stages; nesting behaviour may maximize developmental success of eggs, but not necessary maximize the survival of mother and hatchling. For example, nesting in open areas may provide suitable thermal environments for successful embryonic development, but could incur a higher risk of predation on nesting mothers. Embryos and hatchlings may have different optimal temperatures [183], implying differential effects on embryos and hatchlings of nests in terms of location or depth. In box turtles (Terrapene carolina carolina, Clemmys guttata), deeper nests increase hatching success whereas shallower nests favour juvenile survival to the overwintering stage [184]. Maternal nest-site choice may thus balance opposing pressures on mother, egg and hatchling to optimize reproductive success. This scenario suggests that when evaluating the adaptive significance of nesting behaviour we should consider the impact of nesting behaviour not only on embryonic development, but also on maternal and hatchling fitness.

Third, although adaptive nesting behaviours have been reported in a number of species in response to environmental change, the evolutionary potential of this behaviour has not been explicitly demonstrated. Insights into this topic could not only provide insight into fundamental issues about the evolution of maternal effects, but also let us predict whether nesting behaviour can evolve at a sufficient rate to keep pace with climate warming [48]. In addition, the evolution of nesting behaviour may also affect natural selection on embryonic plasticity in response to climate change, because maternal nesting behaviour could buffer embryos from environmental variation [47]. Currently, we know that nesting behaviour exhibits significant spatio-temporal variation in reptiles (see §2), but whether these variations are due to phenotypic plasticity or are heritable (genetically based) is unexplored. If such variations possess additive genetic effects, we may expect adaptation of nesting behaviour to climate change, so long as the pace of environmental change does not exceed the potential rate of evolutionary change [46]. However, current evidence suggests that nest-site behaviour has little or no heritability, complicating any predictions about the ability of buffering mechanisms to evolve [46]. Studies on the heritability of nest-site selection behaviour in a wide range of reptile species are needed to fill this gap in our understanding of the evolution and adaptive potential of nesting behaviour.

Data accessibility

The data are provided in electronic supplementary material [185].

Authors' contributions

R.S.: conceptualization, writing—review and editing; W.-G.D.: conceptualization, formal analysis, investigation, methodology, project administration, supervision, writing—original draft, writing—review and editing; B.-J.S.: formal analysis, investigation, methodology, writing—original draft; S.-R.L.: investigation, writing—original draft.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This work was supported by grants from National Natural Science Fund of China (31821001, 32171486) and The Strategic Priority Research Program of the Chinese Academy of Sciences (XDB31000000).