The research team, led by scientists at the University of Brasília and partner institutions across Brazil and the United States, set out to address a persistent blind spot in climate change biology. Most studies of warming impacts on reptiles focus on diurnal species, even though more than a quarter of the world’s lizards are nocturnal, primarily within the gecko lineage Gekkota. Climate projections often treat nighttime warming as negligible or assume temperatures rise uniformly across the day, leaving the thermal worlds of night-active animals poorly characterized. Because nocturnal species cannot bask in sunlight to adjust their body temperatures, they depend heavily on the microclimates available in refuges and on the thermal properties of the substrates they occupy, making them uniquely sensitive to how nights and refuges warm.

To build their assessment, the researchers captured 43 individual geckos in the Bacaba Municipal Park in Nova Xavantina, Mato Grosso, in the western Cerrado near the ecotone with Amazonia. Sampling took place at twilight and during the early night, coinciding with the species’ natural activity window. In the laboratory, they measured four suites of physiological traits: preferred body temperatures, voluntary thermal limits, locomotor performance, and critical thermal tolerance. Lizards ran along a two-meter track filmed at 420 frames per second, allowing the team to extract maximum sprint speeds at different body temperatures. Preferred temperatures ranged from 16.3 to 39.7 degrees Celsius, with the fifth and ninety-fifth percentiles falling at 17.8 and 34.5 degrees, while the critical thermal minimum and maximum were measured at 7.9 and 40.5 degrees respectively. Locomotor performance peaked at 31.6 degrees, and the best-fitting thermal performance curve included body size as a covariate, confirming that larger animals moved differently across the thermal landscape.

These physiological measurements fed into three parallel modeling strategies. The first was a purely correlative species distribution model, relating 42 filtered occurrence records to macroclimatic variables such as mean diurnal range, isothermality, and precipitation seasonality drawn from the WorldClim 2 database at 2.5-kilometer resolution. The second was a mechanistic model built exclusively from ecophysiological predictors, including rasters of hours of activity and locomotor performance derived from the field data. The third was a hybrid model combining both sets of predictors. All models were fitted with the random forest algorithm using down-sampled pseudo-absences and evaluated with four-fold cross-validation, with final predictions generated as weighted ensembles of models scoring above 0.7 in true skill statistic. When the three approaches were compared using linear mixed-effects models, the hybrid model performed slightly better, and the researchers adopted it as their final projection tool.

The hybrid model’s projections under three shared socioeconomic pathway scenarios, from the optimistic SSP245 to the pessimistic SSP585, using the EC-Earth3-Veg global circulation model, painted a uniformly grim picture. By the period 2081 to 2100, the species is projected to lose between 96.2 and 97.3 percent of its climatically suitable range, with the current range of roughly 63,800 pixels shrinking to fewer than 2,500 stable pixels in every scenario. Even allowing for dispersal into newly suitable lead areas, future range sizes remain below 4,000 pixels, representing overall range changes of between 93.9 and 96.3 percent. The most suitable remaining areas concentrate in the central Cerrado and a portion of the southern Amazonia-Cerrado transition, but these refugia are small and fragmented. Mean diurnal range, isothermality, and precipitation seasonality emerged as the most important predictors, while ecophysiological variables contributed additional but not dominant explanatory power, suggesting that thermal physiology refines rather than replaces macroclimatic signals.

Range contraction, however, tells only half the story. The team also deployed biophysical simulations using the NicheMapR package to model how microclimates and lizard activity would change at 900 randomly sampled coordinates across the rear, stable, and lead areas. The simulations computed hourly microclimates for a full year under open-surface conditions, representing the most thermally extreme environments available to the gecko, and then calculated heat and water exchange between the animal and its surroundings using morphological, physiological, and behavioral traits. Under present conditions, the species enjoys approximately 12,500 active hours per year, when environmental temperatures fall within its voluntary thermal limits. Under all three warming scenarios, that figure drops to fewer than 5,000 hours, a decline of more than half, even though the animals will need to spend less time surface heating as the environment warms.

That reduction in surface heating time might at first seem like good news, since lizards basking in the open are more exposed to predators. But the trade-off turns sour: fewer total active hours means less time for foraging and reproduction, which cascades into population-level consequences. The simulations also revealed a shift in the nature of thermal risk. Under present conditions, no air, substrate, or body temperatures ever exceeded the critical thermal maximum, but future warming scenarios produced exceedances, indicating that acute heat exposure will become a genuine hazard. Cold stress, by contrast, is unlikely to ever limit the species, as temperatures rarely if ever fell below the critical thermal minimum. The researchers also calculated degree-hours, which integrate both the intensity and duration of thermal stress, and found no accumulation of sublethal stress under current conditions, though the picture changes as warming intensifies.

One of the study’s most technically interesting findings concerns rock-air decoupling, a measure of how much rock microclimates diverge from ambient air temperatures. The decoupling values were consistently positive across all scenarios, meaning rocks act as thermal capacitors, absorbing heat during the day and releasing it slowly at night. This buffering reduces the risk of sudden cooling and extends opportunities for nocturnal and crepuscular activity. Crucially, both the mean and the ninety-fifth percentile intensities of decoupling remained stable across all warming scenarios, suggesting that the functional role of rocks as thermal refugia will persist even as maximum and minimum environmental temperatures rise. However, the researchers caution that even slight increases in minimum night temperature could reduce the thermal gradient that rocks provide, and the buffering effect, while persistent, is not powerful enough to counterbalance the dramatic loss of activity time and geographic range.

The findings place Gymnodactylus amarali in a distinctly more precarious position than many of its relatives. Other studies have found that populations of the Andean lizard genus Stenocercus will face thermal deficits but retain refugia that prevent performance loss, that the Amazonian lizard Kentropyx calcarata faces extinction risks of 26.1 to 70.8 percent by 2050, and that the South American genus Teius shows only minor declines, while the desert lizard Liolaemus ditadai may actually gain suitable habitat. The contrast underscores that climate vulnerability is species- and context-specific, shaped by physiology, behavior, and microhabitat availability. For tropical ectotherms generally, which already live close to their upper thermal limits, the margins for error are narrow, and the sharp decline in locomotor performance above optimal temperatures observed in this gecko mirrors patterns documented across other tropical lizard species.

The authors emphasize that their simulations did not explicitly incorporate changes in precipitation regimes, drought intensity, or land-use dynamics, all of which could substantially alter local microclimatic conditions in the Cerrado. Vegetation removal associated with agricultural conversion can increase thermal exposure and reduce microhabitat buffering, while irrigation may locally mitigate thermal stress by increasing moisture availability. These contrasting effects mean that the microclimates actually experienced by individuals may deviate from broad-scale predictions, particularly in human-modified landscapes. The study’s conservation message is nonetheless clear: protecting rocky habitats and remaining Cerrado remnants that maintain thermal heterogeneity is critical, not just for this gecko but for the broader community of saxicolous species that depend on these structures. As nocturnal tropical lizards remain underrepresented in the climate vulnerability literature compared to their diurnal counterparts, this integrative approach, fusing physiology, distribution modeling, and biophysical simulation, offers a template for assessing the hidden risks facing the night-shift members of tropical ecosystems.

Subject of Research: Climate change vulnerability of the nocturnal saxicolous gecko Gymnodactylus amarali in the Brazilian Cerrado, assessed through hybrid species distribution models and biophysical simulations

Article Title: Effects of climate change on a nocturnal gecko: an integrative analysis

Article References: Costa, T. D. L., Serpa, A. D. O., Sena, A., de Souza, B. A., de Oliveira Caetano, G. H., de Sousa, H. C., da Silva, J. M. A., Godinho, L. B., Diele-Viegas, L. M., Silva, N. A., Cavalcante, V. H., & Colli, G. R. (2026). Effects of climate change on a nocturnal gecko: an integrative analysis. Frontiers in Amphibian and Reptile Science, 4, Article 1753497. https://doi.org/10.3389/famrs.2026.1753497

Image Credits: AI Generated

DOI: 10.3389/famrs.2026.1753497

Keywords: climate change, Gymnodactylus amarali, Cerrado, nocturnal gecko, species distribution models, biophysical modeling, thermal physiology, ectotherms, habitat loss, rock outcrops, microclimate, conservation