Sacred Fir Survivors Show Subtle Genomic Signs of Selection

Sacred fir seedlings that survived a hot, dry planting at the lower edge of their natural range carried subtle genetic signatures consistent with selection, according to a genomic study of an assisted-migration experiment in central Mexico. Researchers compared 19 survivors from the 2,600-m elevation garden with 216 seedlings from higher gardens and identified four genomic loci that appeared both potentially selected and associated with source-climate variation. The results suggest that heat- and drought-exposed survivors may have differed genetically from the original group, although the limited functional annotation and small number of survivors make the biological meaning of individual variants difficult to establish.

The experiment was established in 2019 in the Monarch Butterfly Biosphere Reserve, which spans the states of México and Michoacán. Seedlings from 11 sacred fir, Abies religiosa, provenances between 3,003 and 3,552 m above sea level were planted in common gardens at 2,600, 3,000 and 3,400 m. Each garden contained 330 seedlings in six blocks, with 30 seedlings from each provenance. By early 2022, more than 84% of seedlings from every provenance remained alive at the middle garden and more than 90% remained alive at the upper garden, whereas survival at the lowest garden ranged from zero to 13.3% for individual provenances.

That severe mortality created a natural filter through which researchers could examine the genetics of the remaining seedlings. A shift toward warmer and drier conditions is already associated with increased sacred fir mortality, including deaths linked to hotter droughts and bark beetle outbreaks. Modeling cited by researchers projected that the area suitable for the species nationwide could shrink by 96% by 2090, while the climate-sensitive portion of the Monarch Butterfly Biosphere Reserve was among the areas expected to be severely affected. The reserve also has ecological importance because dense sacred fir crowns provide migrating monarch butterflies with a moderate microclimate.

At the low-elevation site, the relationship between survival and the distance a provenance was moved was positive but not statistically significant, with a generalized linear model yielding p = 0.180. No significant transfer-distance relationship was detected at 3,000 m (p = 0.537) or 3,400 m (p = 0.311). In other words, seedlings from sources closest to the warm range edge tended to survive somewhat better at 2,600 m, but elevation displacement alone did not adequately explain survival within any garden. The investigators therefore turned to genome-wide markers to look for differentiation and climate associations that a simple elevation measure might miss.

Needle samples from all 19 survivors in the low garden and 17–20 individuals per provenance from the middle and high gardens were analyzed using genotyping-by-sequencing, which produced 1,511,227 variant sites before stringent filtering. Quality control retained 18,816 high-quality single-nucleotide polymorphisms, or SNPs, including 11,780 transitions and 7,036 transversions. The researchers used a de novo assembly and SNP-calling pipeline because the only available Abies genome sequence belonged to European silver fir, a relatively distant relative of sacred fir. The comparison group combined the 216 seedlings from the middle and high gardens, where survival was high, as an estimate of the allele frequencies present before selection at the low site.

Overall genetic differentiation between the initial and survivor groups was very low, with an F_ST of 0.00256 across the retained SNPs. Principal components explained small proportions of the total variation, and source provenances were not completely homogenized: some clustered more distinctly than others, with the most differentiated sources located at 3,552, 3,457, 3,364 and 3,300 m. Because the survivors were a subset of the original group, strong separation across the entire genome was not expected. Instead, the researchers searched for loci that departed from neutral population structure and separately tested whether alleles covaried with climate variables at their sources.

PCAdapt identified 2,202 outlier SNPs with patterns that could be consistent with selection. A genotype-environment analysis using latent factor mixed models identified 33 SNPs associated with five climate variables: degree-days below 0°C, July–August precipitation, November–February precipitation, the Julian date of the last spring freeze, and the ratio of summer to spring precipitation. One SNP was associated with all five variables, six with more than one and 26 with only one. Winter precipitation accounted for 15 associations, summer precipitation for nine, cold degree-days and the last spring freeze for 11 each, while summer precipitation balance accounted for two. Only four of the 33 climate-associated SNPs overlapped with the PCAdapt outliers; three were linked to winter precipitation and one to degree-days below 0°C.

Minor-allele frequencies among the low-garden survivors were more variable than in the initial group for both sets of candidate loci. Mean minor-allele frequency was 0.127 ± 0.104 among survivors versus 0.124 ± 0.089 in the initial group for the PCAdapt-selected SNPs; for the climate-associated SNPs, the respective values were 0.133 ± 0.139 and 0.137 ± 0.100. A formal F-test did not find a statistically significant difference in variance between the groups for the 33 climate-associated loci. Nevertheless, cumulative differentiation across loci suggested possible adaptive differences, with a few loci approaching moderate effect sizes of 0.40–0.46. The evidence therefore points to small, distributed genomic shifts rather than a single clearly identified “survival gene.”

Several limitations constrain the interpretation. Only 19 seedlings survived at the hot lower garden, creating a sharply unbalanced comparison with 216 seedlings in the initial group. The genotyping-by-sequencing data also had high error and missing-data rates, and low linkage disequilibrium in conifers means that a sparse set of SNPs may miss nearby variants responsible for local adaptation. Family structure or relatedness among seedlings could also influence the results, while the linear climate models used in the analysis may not capture nonlinear responses to drought or other extremes. In addition, the 11 source populations came from a limited part of the species’ range and did not represent its full climatic diversity. Soil conditions, which may also affect sacred fir genetic responses, were not included in the climate association model.

The findings provide a promising lead for future field testing rather than a ready-made tool for selecting drought-tolerant trees. If alleles linked to low winter precipitation or cold degree-days improve performance in particular environments, they could eventually inform genomic prediction or assisted-migration programs. Researchers would still need to test candidate loci in larger planting experiments, develop genomic and transcriptomic resources for sacred fir and determine whether the observed variants lie within or near functional genes. The study’s central result is that the unusually harsh 2,600-m garden exposed pre-existing genetic variation within sacred fir: surviving seedlings did not look dramatically different overall, but a small set of loci showed patterns consistent with climate-related selection and could guide more definitive experiments on how this forest tree may persist as its suitable range contracts.

Subject of Research: Genetic evidence of selection in heat-exposed sacred fir seedlings from a Mexican range-edge planting experiment

Article Title: Do Abies religiosa Seedlings That Survived Hot/Dry Range‐Edge Conditions Exhibit Genetic Evidence of Selection?

Article References: Du, Y., Moran, E. V., Zhang, R., Cruzado‐Vargas, A. L., & Sáenz‐Romero, C. (2026). Do Abies religiosa Seedlings That Survived Hot/Dry Range‐Edge Conditions Exhibit Genetic Evidence of Selection?. Ecology and Evolution, 16(10), Article e74455. https://doi.org/10.1002/ece3.74455

Image Credits: AI Generated

DOI: 10.1002/ece3.74455

Keywords: sacred fir, Abies religiosa, genomic selection, climate adaptation, assisted migration, SNP analysis, seedling survival, Mexico, range-edge ecology, forest conservation, Abies, religiosa