1 INTRODUCTION

Impacts of climate change are manifested across the tree of life, but can be particularly potent for ectotherms, which make up the majority of animal biodiversity (Wilson, 1992). For these taxa, environmental temperature determines metabolic rate, which governs patterns of growth, reproduction, locomotion efficiency, and immune responses (Schmidt-Nielsen, 1997). For many ectothermic species, climate change is already profoundly impacting population dynamics, ecology, behavior, and morphology (Deutsch et al., 2008, 2015; Gunderson & Leal, 2016; Kearney et al., 2009; Kingsolver et al., 2013; McCauley et al., 2018). For some ectothermic taxa, including many reptiles and teleost fish, rapidly changing environmental temperatures may influence sex determination during development, and therefore sex ratios of populations (Janzen, 1994; Mitchell & Janzen, 2010; The Tree of Sex, 2014). Temperature-dependent sex determination (TSD) occurs in some lizards, all crocodilians, and most turtle species (Janzen & Paukstis, 1991). Three patterns are recognized based on one or two “pivotal temperatures,” or those temperatures that give approximately equal numbers of males and females, (Bull, 1980): (A), in which high and (B) low incubation temperatures result in females; and (C), with two pivotal temperatures, in which low and high temperatures produce females, and intermediate temperatures produce males. Temperature-dependent sex determination may have evolved in lineages in which temperature during development influences male and female fitness differently (Charnov & Bull, 1977). The three reptile lineages (Lepidosauria, chelonian, crocodiles) which display TSD differentiated from a common ancestor over 200 million years ago. Since this time, mean global temperatures have fluctuated considerably (10–20°C) and over a few decades during the Holocene yet many lineages with TSD survived (Mitchell & Janzen, 2010). Rapid temperature fluctuation over geological time was limited mostly to the northern hemisphere, which may have allowed some taxa to persist (Mitchell & Janzen, 2010). Others survived changing thermal conditions through adaptation and/or phenotypic plasticity (Refsnider & Janzen, 2016). Current climate change is comparable in rate (Dt = 0.2–0.6°C per decade) and magnitude (Dt by 2100 = 1.3–5.7°C above current mean) (Wuebbles et al., 2017) but is global in scale. What is not comparable is the conservation status of species and populations that display TSD during the current rapid phase of global warming (Figure 1).

The Chelonia (turtles and tortoises) is the most imperiled vertebrate order, with more than half (61%) of the 356 known species threatened with extinction due to over-harvesting (e.g., illegal trade), infectious diseases, habitat destruction, and climate change (Lovich et al., 2018; Rhodin et al., 2021). Of the 79 chelonian species tested, over 80% display TSD (Ewert et al., 2004). Species with TSD are composed of small, endangered populations that often face multiple threats, with low genetic and phenotypic diversity, may lack the resilience to withstand the impact of rapid temperature change on sex ratios, effective population size, or emergence time/food availability mismatch (Mitchell & Janzen, 2010). A growing body of empirical and modeling evidence demonstrates that nest temperatures of marine and freshwater turtles are increasing under climate change and skewing sex ratios (Fuentes & Porter, 2013; Hawkes et al., 2009; Jensen et al., 2018; Patrício et al., 2021; Refsnider et al., 2013; Santidrián Tomillo et al., 2015). It is likely that terrestrial Chelonians are, or will be, similarly impacted as has been shown in the red-footed tortoise (Hernández-Montoya et al., 2017).

Among chelonians, Galapagos tortoises (Chelonoidis spp.) are conservation icons and emblematic of terrestrial tortoises worldwide. Of the 14 known Galapagos tortoise species, at least 12 remain extant, of which six are critically endangered, three are endangered, and three are vulnerable (IUCN, 2022). Although uncertain, overall predictions for the Galapagos trend to a warmer and wetter climate with both El Niño and La Niña (ENSO) events more frequent and intense (Restrepo et al., 2012; Trueman & d'Ozouville, 2010). Changing temperature and precipitation will have long-term impacts on Galapagos flora and fauna (Charney et al., 2021), with recent evidence of rapid anthropogenic environmental changes in Galapagos, including climate, triggering intense selection pressures for adaptive phenotypes and rapid evolution (Deem et al., 2010; Grant et al., 2017; Grant & Grant, 2003). More generally, the potential for rapid evolution and/or phenotypic plasticity to keep pace with environmental change is a concern globally and across the tree of life as the Anthropocene proceeds (Carroll et al., 2014; Johnson & Munshi-South, 2017). This challenge is perhaps greatest for long-lived, long-generation time endemic species on islands comprised of small, isolated populations that have experienced one or more genetic bottlenecks (Graham et al., 2017). All these traits are exemplified by giant Galapagos tortoises (Jensen et al., 2021; Poulakakis et al., 2012).

Giant tortoises occur on seven of the Galapagos Islands (Caccone et al., 2002) (eight with a recently introduced population to Santa Fe Island) (Tapia et al., 2022). The topography of the Galapagos Islands varies from relatively flat, older islands to young active volcanoes reaching maximum elevation of 1900 m above sea level (ASL) (Tye et al., 2002). Tortoises occur across the entire elevation gradient. Female nest site selection varies by island and within populations and can occur from close to sea level to at least ca. 1000 m ASL (SB pers. obs.). The combination of large seasonal, interannual, and elevational variation in insolation, air temperature, and vegetation cover are likely to have a strong influence on soil temperature and therefore potentially on hatchling sex ratios; however, this has never been documented. Understanding patterns of sex determination among recruits into populations, and the drivers of these patterns is a high research priority (Kubisch & Ibargüengoytía, 2021).

From studies conducted in captivity, Galapagos tortoises are known to display the A pattern of TSD with warmer temperatures producing females and cooler temperatures producing males (Marquez et al., 1990; Sancho et al., 2017). In the Tortoise Reserve on Santa Cruz Island at least six tortoise nesting areas are known to rangers from the Galapagos National Park Service. Within one of these areas, El Chato, three distinct nesting zones occur along an elevational gradient between 10 and 140 m ASL (Figure 2), which are included in a long-term study on Galapagos tortoise movement ecology (Blake et al., 2013; Blake, Yackulic, et al., 2020). Application of a thermal model developed by Blake, Parlin, et al. (2020) indicates a ca. 1°C decline in mean annual ambient air temperature from lower to upper nesting zone and more persistent cloud cover decreases direct solar radiation as elevation increases (Trueman & d'Ozouville, 2010). While air and soil temperatures are correlated (Blake, Parlin, et al., 2020), variation in nest temperatures over the elevation gradient are unknown, which motivated this study.

We hypothesized that the temperature gradient over three nesting zones would result in a higher frequency of female hatchlings produced in warmer, lower elevation nests, and more male hatchlings produced in the cooler, upper elevation nests. To test our hypothesis, we measured nest temperatures and determined the sex of juvenile tortoises residing in the three nesting zones along the elevational gradient. Here we discuss the implications of our results for Galapagos tortoises within the context of projected climate change in the Galapagos archipelago.

2 METHODS

3 RESULTS

3.1 Sex determination

Juveniles were collected at elevations between 10 and 112 m ASL, and temperature data were collected from nests between 6 and 149 m ASL. All juveniles appeared in good physical condition on visual examinations. We successfully identified the sex of 40 animals from the three nesting zones (9, 21, and 10 from the lower, middle, and upper zones, respectively). Body weights ranged from 77 to 1358 g, with a mean of 529 g. Of 9 and 21 juveniles collected and sexed in the lower and middle nesting zones, respectively, only one (11.1%) and two (9.5%) were male, while from the upper nesting zone, eight out of ten (80%) juveniles were male (Figure 3a). In support of our hypothesis, a logistic regression indicated a significant positive relationship between increasing nest elevation and the probability of being male (Deviance Ratio _(2,37) = 8.43, p < .001). Juvenile weight had no significant effect on sex ratios, suggesting that incubation temperature was the primary determinant of sex ratios. Parameter estimates were as follows: Elevation = 0.04 (SE = 0.0167), p = .017; Juvenile weight = −0.00256 (SE = 0.00184), p = .163.

4 DISCUSSION

Consistent with our hypothesis, we demonstrated that (1) nest temperatures of free-living C. porteri were negatively correlated with elevation, and (2) the frequency of male juvenile tortoises in the population was positively correlated with increasing elevation. It is noteworthy that the total elevational change across all nesting zones was little over 100 m, and the switch from ca. 10% males to >80% males occurred between 57 m and 104 m mean nest elevation.

We recorded considerable fluctuation in nest temperature both diurnally and over longer periods, though generally mean nest temperature increased throughout the incubation period. The impacts of these variable thermal regimes on sex determination (and indeed development and survival more generally) are not known for this species. The highly variable thermal regimes of nests in the field and our very small sample sizes preclude any attempt to define a pivotal temperature (Ewert et al., 1994, 2004; Janzen & Paukstis, 1991; Warner & Shine, 2011). Sancho et al. (2017) estimated a pivotal temperature of 28.3°C for captive Española Island tortoises (Chelonoidis hoodensis), based on empirical data in which a consistent incubation temperature of 25.5°C produced 91% male hatchlings, while 33% males and 67% females resulted from an incubation temperature of 29.5°C. The Sancho et al. (2017) model predicted a large “transition range” from 25.2°C to 31.4°C over which both males and females would be produced. A large transition range would provide considerable buffering against potentially catastrophic demographic impacts of climate change (Sancho et al., 2017). However, the relatively small range of mean incubation temperature and dramatic change in sex ratio demonstrated in our study suggests a much narrower transition temperature range.

The difference between the nest temperatures we recorded, and the incubation temperatures used at the Tortoise Breeding Centers on Galapagos (28°C for males, 29.5°C for females) is noteworthy. In our study, mean nest temperatures taken in the middle of the nesting chambers rarely exceeded 28°C, and never surpassed 29.5°C during the middle trimester of incubation (Figure 2b,c). There are no published data on the outcome of these incubation temperatures (i.e., below 28°C and above 29.5°C) from the Captive Breeding Centers: however, our data would suggest that tortoises incubated in captivity may have sex ratios heavily skewed toward females.

We attributed the change in sex ratios across the nesting zones to differences in incubation temperatures. We can rule out differential mortality of males and females as the cause of sex ratio differences observed because mean body size by sex was not significantly different across nesting zones. However, our data do not allow us to rule out sex-related egg mortality differences by zone, though this seems unlikely.

Our study was motivated not only by obtaining the first indication of how nest location may influence nest temperature and sex ratios of offspring but also by the question of how a changing climate may impact Galapagos tortoise sex ratios, and the potential consequences for tortoise demography and conservation efforts. Our study was limited to three nesting zones in a small part of the range of one species, so we are unable to generalize to the population level, species, or beyond across the archipelago. We do not know current sex ratios of populations or species, nor do we know the full distribution of nest sites or their thermal regimes during incubation. However, we can predict that the projected increase in ambient temperature of Galapagos of between 1 and 4°C (Charney et al., 2021) will highly skew the sex ratio toward females. In the short term, more females will likely increase population growth rates (Blechschmidt et al., 2020; Butler, 2019; Deutsch et al., 2008), since the abundance and survival of reproductively active females largely determines population growth rate in long-lived species such as Galapagos tortoises (Gibbs & Goldspiel, 2020). However, a dramatic reduction in male recruitment would further erode genetic diversity and reduce effective population size (Reed & Frankham, 2003) of these already genetically compromised species (Milinkovitch et al., 2004, 2013; Russello et al., 2005). An alternative outcome may be related to predictions of increased precipitation that would stimulate vegetation cover and reduce direct solar radiation onto the soil. Potentially this could counter ambient warming and reduce nest temperatures favoring male hatchlings. However, the projected increases in rainfall is also likely to have a serious negative impact on egg and hatchling mortality, over-riding the effects of sex ratio change in many parts of Galapagos (Charney, 2021).

Galapagos tortoises on some islands nest anywhere between 100 and 1000 m elevation, but we can make no general predictions on sex ratios with elevation from our data because interactions between elevation, aspect, sea surface temperature, and dry season cloud cover generate heterogeneous temperature profiles (Blake, Parlin, et al., 2020). There is a huge geographic range over which Galapagos tortoises could potentially nest to optimize fitness (which includes offspring sex ratios) in a changing climate. However, we do not know to what extent Galapagos tortoises are philopatric to nest sites or whether females can make decisions on where to nest based on current climate and other factors. Studies of chelonians elsewhere suggest that strong philopatry is common (Freedberg et al., 2005; Lee et al., 2007; Stiebens et al., 2013; Walde et al., 2007) which may limit a plastic response to nest site location by females as a response to rapid climate change. If nest site choice does exhibit plasticity, Galapagos tortoise species that live on islands with a large elevation range may be able to shift their nesting sites to higher, cooler, elevations to maintain consistent sex ratios under climate change. In fact, behavioral plasticity and rapid evolutionary response may have potential to mitigate catastrophic climate impacts for some species (Gunderson & Leal, 2016; Kingsolver et al., 2013). Furthermore, plasticity among female chelonians in nest depth has been suggested as a possible mechanism to maintain stable sex ratios in response to climate variability, though conclusions are equivocal (Marco et al., 2018; Ozdemir et al., 2011; Refsnider et al., 2013).

5 CONCLUSION

This is the first study to quantify the potential impact of temperature on sex ratios of Galapagos tortoises in the wild. We demonstrated that a small (ca. 50 m) increase in elevation induced a significant drop in nest temperatures, and we documented a flip in sex ratios of juveniles from ca. 10% males at low elevations to 80% males at the upper elevation. Over the next 30–50 years, mean temperatures on Galapagos are predicted to become hotter than at any time over the last 120,000 years (Charney et al., 2021). This will likely increase the F:M ratio in many nesting zones, but the overall consequences for populations and species of Galapagos tortoises cannot be determined because of several unknown factors including: 1. the distribution of nest sites, 2. the thermal regimes of nests over time and space, 3. pivotal and transitional temperatures, and 4. the range of behavioral plasticity among nesting females. Nonetheless, our study highlights the need to include the potential impact of TSD on Galapagos tortoise demographics in conservation planning and management of these iconic species.

AUTHOR CONTRIBUTIONS

Sharon L. Deem: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); resources (equal); supervision (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Sam Rivera: Formal analysis (equal); investigation (equal); methodology (equal); resources (equal); writing – review and editing (equal). Ainoa Nieto-Claudin: Investigation (equal); methodology (equal); project administration (equal); writing – review and editing (equal). Evan Emmel: Investigation (equal); methodology (equal); writing – review and editing (equal). Freddy Cabrera: Investigation (equal); methodology (equal). Stephen Blake: Conceptualization (equal); data curation (equal); formal analysis (equal); funding acquisition (equal); investigation (equal); methodology (equal); project administration (equal); resources (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal).

FUNDING INFORMATION

This work was supported through Saint Louis Zoo Institute for Conservation Medicine, the Galapagos Conservation Trust, Houston Zoo, The Swiss Friends of Galapagos, the National Geographic Society Committee for Research and Exploration, and the National Science Foundation (DEB 1258062). VSN International (Hemel Hempstead, UK) provided free statistical software.

CONFLICT OF INTEREST STATEMENT

We declare that we have no competing interests.