A reproducible artificial-life world for ecological perturbation, inherited agent state, and genotype-derived digital assets

We introduce CHIMERIA Digital Aquarium v0.1.0, an open computational research environment in which locally acting software organisms move, consume finite resources, interact as grazers and hunters, reproduce, transmit mutable parameters, and produce inspectable descendants. Each organism has persistent identity and a lineage. A bounded energy ledger tracks externally introduced nutrients and modeled dissipation. We present a controlled in-silico intervention study comprising 36 paired random seeds, five experimental arms, and 160 ticks per run. At the end of the intervention window, the reference arm averaged 25.97 grazers, resource deprivation averaged 2.47, and combined oxygen/toxin/temperature stress averaged 0.67. The corresponding paired mean differences were -23.50 and -25.31 organisms, respectively. Under combined stress, grazer extinction occurred in 61.1% of worlds; after the environment was re-stored, the median remained zero despite a mean of 21.33, demonstrating strong between-world heterogeneity. A nonheritable parameter control did not establish a significant adaptive advantage for inheritance under these conditions. Digital specimens also export deterministic vector artwork and 3D meshes derived from their own genome parameters. The approach is auditablebut is not a biologically calibrated virtual cell, a wet-lab organism, or evidence of consciousness. We present implementation details, negative and positive controls, numerical invariants, raw artifacts, threats to validity, and a staged plan toward multiscale and experimentally constrained artificial-life systems