Aim: The impact of climate change on biodiversity is often analyzed under a stable evolutionary perspective focused on whether species can currently tolerate warmer climates. However, species may adapt to changes, and particularly under conditions of limited fragmentation, adaptive genetic variation can spread across populations tracking changing climates, a process termed evolutionary rescue.  Here, our aim is to integrate genomic data, niche modeling, and landscape ecology to predict range shifts and the potential for evolutionary rescue.

Location: the megadiverse Amazonian rainforest

Methods: We use genome-environment association analyses to search for candidate loci under environmental selection, while accounting for neutral genetic variation in a widespread Amazonian whiptail lizard (Teiidae: Kentropyx calcarata). We then model the distribution of individuals with genotypes adapted to different climate conditions. Finally, we predict range shifts in distinct future climate change scenarios by integrating this information with dispersal constraints based on predicted scenarios of forest cover across Amazonia.

Results: We find that the potential for evolutionary rescue and, therefore, a smaller degree of range loss buffering extinction risk in the future is considerably high, provided that current forest cover is retained and climate change is not extreme. However, under extreme environmental change scenarios, range loss will be high in central and southern Amazonia, irrespective of the degree of deforestation.

Main Conclusions: Our results suggest that protecting the Amazonian rainforest against further deforestation and mitigating climate change to moderate scenarios until 2070 could foster evolutionary rescue. These actions could prevent substantial biodiversity loss in Amazonia, emphasizing the importance of understanding species adaptability in maintaining biodiversity.

We sampled individuals of K. calcarata in Amazonia, the Atlantic Forest, and the Cerrado savannas (information on specimens, museums, and localities are available in Supporting Information, Table S1).

We extracted genomic DNA from each sample (liver and muscles) using the Macherey-Nagel^® Mini Kit with a high salinity protocol with proteinase K. We then visualized fragment sizes in agarose gels and measured DNA concentration and quality with Qubit™ 3.0 Fluorometer and Nanodrop, selecting only samples with a low degree of DNA fragmentation and normalizing the concentration to 20 ± 2 ng DNA for each 50 μl solution (TE buffer). RAD-sequence library preparation (as detailed in Etter et al., 2011) was then performed by Floragenex, Inc. (Eugene, OR, USA). Individual samples were digested with the SbfI enzyme, linked with barcoded RAD ligators, multiplexed, sonicated, and then size-selected to a range within 300-500 bp. After PCR amplification, DNA sequencing was performed on a 2x100bp Illumina HiSeq platform. This procedure resulted in ~8,400,000 reads per individual with a length of 100 base pairs. We deposited the demultiplexed raw-sequencing data in the Sequence Read Archive (details to be provided upon acceptance). 

We demultiplexed and assigned the reads for each sample and performed a de novo assembly and SNP calling using iPyrad 0.9.55 (Eaton & Overcast, 2020), checking the quality of demultiplexed samples using MultiQC (Ewels et al., 2016). For the iPyrad pipeline, we allowed one mismatch from individual barcodes, clustering the reads of each sample (minimum size of 35 base pairs) and across the samples (de novo assembly), sequence coverage of 6x, and minimum Phred score of 33. We selected a minimum clustering threshold of 0.90 after testing values from 0.85 to 0.95 for missingness and number of recovered SNPs (McCartney-Melstad et al., 2019). We allowed a maximum of two alleles and a maximum proportion of 0.5 for heterozygous sites per locus, including only loci present in 75% of the individuals.