Table 1 Applications of genomics data to relevant steps in the decision framework

Can species tolerate change in situ?

Determining if a species is currently experiencing stress which suggests it is approaching the limit of physiological tolerance

Screen biomarkers indicative of stress. See [81]

[TS] with [IG]

Gene expression analyses to identify abundance of key gene transcripts

For many species, further research is required to identify biomarkers; however genomics could facilitate this process. A challenge is that biomarkers need to be diagnostic of stress and reproducible—particularly for gene expression markers

Testing whether a species has sufficient phenotypic plasticity to tolerate projected change

Understand the limits to plasticity under environmental change. See [82,83]

[TS] with [IG]

Gene expression analyses. Gene transcript abundance can be used as a surrogate for overarching phenotypic responses

Observing phenotypes will be more appropriate/cost-effective in some cases, but in other cases, gene expression could screen many phenotypes simultaneously at lower cost per sample. For the latter, links to phenotypic data are required

Assessing a species’ historical demography to see how it responded to past climate change

Greater numbers of loci provide the opportunity to reconstruct demographic history deeper in time. See [84-86]

[WGS], [RRS], [DE] with [IG] or [WGS], [DE] with [PPS]

Bayesian skyline plots/coalescent simulations or likelihood-based diffusion modelling from SNP data

Genomics can provide a comprehensive assessment; however, a similar outcome might be achieved using non-genomic tools (e.g. SSRs), particularly where data sets are already available

Do populations have enough genetic diversity for an evolutionary response?

Determining whether the species or population is currently experiencing inbreeding, which can lead to loss of genetic diversity essential for evolution

Genome-wide sequencing allows accurate estimation of heterozygosity in individuals and populations. See [87]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Estimate F-statistics and heterozygosity from SNP data

Non-genomic tools (e.g. SSRs) can be applied to estimate diversity, particularly where data sets are already available. However genomics offers better resolution and diversity estimation. The effects of different levels of diversity on adaptability needs to be established through phenotypic comparisons

Assess whether there is enough standing genetic diversity to provide opportunities to adapt

Accurately estimate the levels of genetic diversity in populations. See [88]

Estimate heterozygosity, DNA sequence diversity estimates (pi, theta) from SNP or sequence data

Determining whether selection has acted on genetic variation in the species

Test whether major events have resulted in selection on genetic diversity. See [89,90]

Allele frequency spectrum tests (e.g. Tajima’s D), linkage disequilibrium, non-synonymous to synonymous polymorphism ratios (e.g. Kn/Ks) from sequence data

Is genetic diversity strongly distributed across populations?

Identification of centres of genetic diversity, or genetically distinct regions, for prioritised conservation

Examine patterns of population genetic structure to identify outlier populations. See [91]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Estimation of population differentiation based on SNP data using classical F-statistics, PCA or MCMC and Bayesian derived estimates of admixture (e.g. STRUCTURE)

Non-genomic tools (e.g. SSRs) could be applied to assess population differentiation. However genomics offers better resolution and accuracy of diversity patterns, which may be important for detecting fine scale structure

Are some populations adapted to local climate?

Identifying whether populations show adaptation to local climate (or other environmental variables)

Identify loci that have been under selection in populations conditioned on local environment. See [92-95]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS], [DE] with [PPS]

Population level or landscape genomics methods based on SNP data: outlier tests, relative rate tests, allelic association with environment and allelic association with adaptive traits. Computational modelling of genomic diversity evolution under environmental change

How to confidently link climate variables to local adaptation, and how to infer adaptive capacity from genomic data are currently pressing questions in population genomics. However these questions also apply to other genetic methods. Characterising the local climate experienced by a species is also a challenge, but microclimate modelling is improving rapidly. Phenotypic data is still essential to determine the extent of adaptation

How quickly can genetic adaptation occur?

Identify rates of genetic adaptation to environment by screening adaptive variation in natural populations experiencing environmental change, or through simulated or experimental evolution. See [96-98]

Is gene flow high enough? (or too high?)

Determining the extent of gene flow between existing populations to inform on dispersal capability and also potential for adaptive alleles to spread or be swamped

Provide estimates of ongoing gene-flow and admixture among populations. See [99,100]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Coalescent genealogy sampling to generate Bayesian and maximum likelihood estimates of migration and gene flow (e.g. Lamarc, Migrate), or MCMC and Bayesian-derived estimates of admixture (e.g. STRUCTURE) based on SNP data. Genomics also has the power to identify recent migrants and so test the efficacy of movement pathways

Genomics can provide a more comprehensive assessment of gene flow compared to non-genomic tools (e.g. SSRs), particularly where rates of gene flow are low

Is a positive evolutionary response possible through natural hybridisation with sympatric species?

Determining whether hybridisation occurs in nature

Estimate rates and genomic extent of hybridisation/gene flow between species in situ. See [101-103]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Identify hybrid ancestry via comparison to known non-hybrids. Estimate migration and gene flow (e.g. Lamarc, Migrate) and admixture (e.g. STRUCTURE) between species. Admixture quantification also confirms F1 hybrid fecundity

Non-genomic tools are available for identifying hybrids; however genomics gives unprecedented power to detect even low levels of introgression, and to understand how patterns of introgression vary across the genome. Phenotypic data are essential to determine whether hybridisation is adaptive

Assess how quickly beneficial alleles can move into a population or species

Track introgression of genomic regions under selection following documented hybridisation events. Identify potential for hybrid incompatibilities or swamping. See [104,105]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Track distribution of species specific alleles in population with regard to null selection models. Transmission distortion in artificial F2 hybrids can indicate genetic incompatibilities

Can species migrate quickly enough?

Assess potential for migration into climatic refugia given ecological constraints and known rates of gene flow

Provide accurate estimates of gene flow (as described above). See [106]

[WGS], [RRS], [DE], [TS] with [IG] or [WGS],[DE] with [PPS]

Genomic estimates of gene flow can be coupled with data on rates of dispersal or movement and habitat analysis (path analyses, resistance models) to predict viability of dispersal pathways

Genomics can provide a more comprehensive assessment of gene flow compared to non-genomic tools (e.g. SSRs), particularly where rates of gene flow are low