Pre-registration: ecology → P1 fraction, out-of-sample panel

Pre-registration of a prospective seven-species test of Perez’s strict-octave prediction λ_Prot/λ_DNA = log 2/log φ ≈ 1.4404 in basal land-plant mitochondria. The hypothesis under test is that the strict-octave regime is the ancestral state of land-plant mitochondrial DNA–proteome coupling, preserved in bryophytes and lycophytes and progressively lost across the radiation of seed plants. Two prior anchor observations (Physcomitrium patens, Selaginella moellendorffii; both sub-1% from the predicted ratio) motivate the test; this pre-registration extends to seven additional species spanning mosses, liverworts, hornworts, and lycophytes. The decision rule is fixed in advance: ≥4/7 sub-1% matches confirms the ancestral-state hypothesis, ≤1/7 refutes it, intermediate values are inconclusive. Six methodological cross-checks (randomised-encoding null, nuclear-genome control, length-filter robustness, codon-bias correction, bootstrap CI, non-basal outgroup) and three robustness conditions are pre-specified. The analysis pipeline has not been executed at the time of pre-registration. Timestamped 2026-05-26.Pre-registration and execution of an out-of-sample test of the c ↔ f_P1 regression bridging the dynamical Lyapunov invariant and the static P1/P2 partition balance of the Master Code of Biology framework. Salinibacter ruber, a halophilic bacterium with reported disorder factor c = 0.7369, was selected as a discriminating test case. The pre-registration fixed the predicted f_P1 = 56.59% with band [54.59, 58.59] before any computational measurement was performed. Subsequent UniProt Swiss-Prot 2026_01 measurement returned 54.75% (strain M31) or 56.29% (species), both within the pre-registered band. The result supports the linearity of the c ↔ f_P1 relationship under composition-based predictions. Pre-registration timestamped 2026-05-26T14:58:38Z; measurement performed in the same session, after the pre-registration was finalised. Pre-registration and execution of an ecology-based out-of-sample test of f_P1 partition balance across a 16-organism panel spanning four ecological strategies (halophiles, hyperthermophiles, mesophiles, psychrophiles) and two domains (bacteria, archaea). Anchored predictions derived from in-sample organisms were stated in advance: halophiles f_P1 ≈ 58 ± 2%, hyperthermophiles ≈ 45 ± 2%, mesophiles ≈ 51 ± 2%, with psychrophiles left as exploratory. Of ten anchored organisms tested, nine fell within the pre-registered band. The single outlier was Chromohalobacter salexigens, a halophilic bacterium using a compatible-solute strategy distinct from the salt-in strategy of the haloarchaeal anchors; this is biologically interpretable rather than a generic failure of the ecology framing. Psychrophiles clustered at 50.7–52.8%, similar to mesophiles — a new exploratory finding. Pre-registration timestamped 2026-05-26T15:30:00Z; measurement performed in the same session, after the pre-registration was finalised.