This instantiation yields a tractable yet biologically realistic model. By focusing on a small set of interdependent traits, it highlights how coordinated adaptation is required for the stooping behavior to function, while remaining mathematically manageable for analysis and simulation.
C.1. Simulation design and parameterization
1. Objectives of the simulation suite
a. Demonstrate whether coordinated multi-trait adaptations (the "peregrine phenotype") can emerge from a CAS model that includes epistasis, pleiotropy, and eco-evolutionary feedback.
b. Identify conditions (parameter regimes) that favor rapid, coordinated evolution versus slow/partial adaptation.
c. Evaluate the role of (a) epistasis/pleiotropy, (b) demographic bottlenecks, (c) prey coevolution, and (d) landscape ruggedness (NK) in producing punctuated vs. gradual dynamics.
d. Produce reproducible outputs (allele trajectories, trait distributions, fitness landscapes, LD/co-selection signatures) for empirical comparison.
2. Model variants (hierarchy)
We propose running a structured ensemble of model variants. Each variant is a fully specified combination of the genetic architecture and ecological coupling:
Variant A --- Baseline (Additive, No Coevolution)
Epistasis Ekj=0E_{kj\ell} = 0Ekj=0.
Pleiotropy minimal (sparse MkM_{k\ell}Mk).
Prey trait distribution fixed (no evolution).
Purpose: baseline behavior (gradual adaptation expected).
Variant B --- Epistasis & Pleiotropy (No Coevolution)
Nonzero epistatic tensor EEE; pleiotropic matrix MMM dense.
Prey fixed.
Purpose: test whether internal genetic interactions alone can produce coordination.
Variant C --- Full CAS (Epistasis, Pleiotropy, Prey Coevolution)
As Variant B plus prey undergo replicator-mutator dynamics with trait-dependent fitness.
Purpose: demonstrate eco-evolutionary feedback (Red Queen).
Variant D --- Bottleneck Scenarios
Same as C but include demographic bottlenecks at specified times (e.g., 0.1 NNN for 5--20 generations).
Purpose: test acceleration of coordinated fixation.
Variant E --- NK Ruggedness Sweep
Implement NK-style genotypefitness (vary KKK across runs: 0,2,4,6).
Purpose: measure effect of landscape ruggedness on punctuated shifts.
Variant F --- Spatial Structure
Add 2D grid with local interactions / migration (dispersal rate mmm).
Purpose: test role of spatial heterogeneity & local adaptation.
Each variant is run as an ensemble (30 replicates) for statistical robustness.
3. Core parameters and default values
Use this table as defaults for exploratory runs. Each parameter should be varied in sensitivity sweeps.
Notes on parameter choice: choose cost coefficients so trade-offs matter (i.e., improvements in one trait are not unconditionally beneficial). Preliminary calibrations should ensure populations neither crash routinely nor trivially saturate.
4. Initialization and random seeds
Genotype initialization: ancestral distribution---e.g., all loci = 0 with low standing variation (add per-locus probability p0 = 0.01 of allele =1), or normal distributed effects for continuous alleles.
Trait baseline: T values computed from mapping; verify that ancestral mean low hunting probability (<0.2) so selection pressure is active.
Ecology baseline: prey trait distribution initialized to moderate evasiveness.
Random seeds: log and fix RNG seeds for every replicate to ensure reproducibility; store seeds in output metadata.
5. Experimental protocols
