1. Predicted genomic signatures and available genomic evidence
Model prediction. Coordinated selection on multi-locus trait bundles produces (i) concerted allele frequency shifts across interacting loci, (ii) extended linkage disequilibrium (LD) / haplotype blocks spanning loci that jointly affect the trait package, and (iii) co-selection signals (covarying selective sweeps) rather than isolated single-locus sweeps. Bottlenecks accelerate fixation and can leave shallow diversity but strong co-selection signatures.
Empirical counterparts & tests.
Candidate loci & pathways. Genes affecting visual system development (opsins, retinal development genes), oxygen transport and respiratory morphology (hemoglobin variants, developmental regulators), neuromuscular function (ion channels, synaptic genes), and feather/wing morphogenesis (keratin/feather genes, ECM regulators) are natural targets for selection scans. Previous comparative avian genomics identifies many of these categories as subject to selection in raptors; targeted analyses in peregrines should test for co-selection.
Genome scans. Composite likelihood ratio tests, extended haplotype homozygosity (EHH), cross-population statistics (XP-EHH, PBS), and site-frequency spectrum tests can detect recent selective sweeps. The CAS prediction is a pattern of multiple nearby or functionally linked signals rather than single isolated peaks.
Linkage & co-selection. High LD or correlated allele-frequency changes among loci (measured by pairwise r2r^2r2 or haplotype block structure) would support concerted sweeps. Time-series genomic data (museum specimens pre-/post-bottleneck, or temporal sampling across populations) would be particularly powerful to detect coordinated frequency shifts.
Comparative genomics & dN/dS. Elevated dN/dS in relevant gene sets across falcon lineages vs non-raptor outgroups can indicate recurrent adaptation; correlated patterns among gene sets (vision + wing + neuromuscular) would match the pleiotropy/epistasis hypothesis.
2. Morphological and developmental evidence
Model prediction. Phenotypic attractors emerge as synchronized trait bundles; transitional forms in the fossil or extant comparative record may display modular (partial) adaptations. Punctuated shifts in morphology are expected where attractor transitions occur.
Empirical counterparts & tests.
Morphometrics. Multivariate morphometric analyses across Falco species and subspecies (wing aspect ratio, wing loading, beak and nares morphology, sternum/keel robustness, retinal area) should reveal trait covariance consistent with coordinated bundles. Strong positive covariation among the focal traits supports the attractor hypothesis.
Ontogeny & evo-devo. Shared developmental pathways (e.g., regulatory genes expressed in both feather and skeletal development) would provide mechanistic bases for pleiotropy. Comparative expression (RNAseq) across developmental stages could reveal co-regulated modules.
Fossil / subfossil record. Although avian fossils are sparse, any temporal sequences showing abrupt morphological shifts (when available) would align with punctuated change predicted by model bifurcations.
3. Ecological and behavioral evidence (Red Queen, trade-offs, bottlenecks)
Model prediction. Coevolutionary coupling with prey produces oscillatory dynamics (Red Queen), shifting adaptive landscapes, and trade-offs that constrain trait optima. Demographic events (e.g., DDT-induced bottlenecks) can accelerate fixation of coordinated alleles or produce transient loss of diversity.
Empirical counterparts & tests.
Prey communities & diet. Quantitative diet studies (stomach contents, prey capture observations, telemetry) that document primary prey taxa and their escape capabilities help parameterize prey trait distributions. Correlations between local prey speed/behavior and peregrine morphology across populations would support eco-selection gradients.
Behavioral ecology & telemetry. High-resolution tracking and high-speed videography quantify stoop kinematics and success rates; shifts in stoop strategy across populations and habitats (urban vs wild) provide direct validation of trait--fitness relationships used in the model.
Historical demographic events. The well-documented DDT bottleneck and subsequent recovery in many peregrine populations offer a natural experiment. Genomic comparisons pre- and post-DDT (or across regions with different histories) can reveal the demographic and selection signatures predicted by the model (reduced diversity, rapid allele frequency changes, possible fixation of co-adapted haplotypes).
Oscillatory dynamics. Long-term ecological datasets (population counts, prey abundance) may reveal cycles consistent with eco-evolutionary feedback; detection of trait oscillations would require paired temporal genomic/phenotypic sampling.
4. Cross-scale, integrative tests (recommended empirical program)
To validate CAS mechanisms rigorously, we recommend a coordinated empirical program that combines:
