osmose.engine.processes.bioen_predation

Bioenergetic predation — allometric ingestion cap. Matches Java BioenPredationMortality.

Unit hazard for whoever writes the bioen overlay. Java’s getMaxPredationRate (BioenPredationMortality.java:241-276) early-returns predationRateBioen[speciesIndex] for background predators (speciesIndex >= nSpecies) before the / getNStepYear() that the focal branch applies. per_fish_ingestion_cap replicates that verbatim, so the FOCAL half of config.bioen_i_max_all is consumed as a per-YEAR rate while the BACKGROUND half is consumed as a per-TIME-STEP rate. Both halves are currently filled from the same per-year key family (config.py: focal predation.ingestion.rate.max.sp{i}, background BackgroundSpeciesInfo.ingestion_rate), so a background Imax authored on the focal convention overshoots by a factor of n_dt_per_year. Background entries must be supplied already in per-time-step units.

A second known gap on the same path: Java reads species.beta.sp{fileindex} for background species (BackgroundSpecies.java:130-133) via cfg.getDouble, which has no default — a missing key is a fatal error("Could not find parameter ...") in Configuration.getParameter (:998-1006). The port has no background beta at all and silently substitutes 0.8, which is the PORT’s focal default (config.py:2499), not a Java default. So where Java refuses to start, the port runs with an invented exponent. Both gaps are Gate-B prerequisites for the overlay, not defects of the cap formula.

Functions

per_fish_ingestion_cap(weight, species_id, ...)

Java getMaxPredationRate x (w*1e6)^beta / subdt, per FISH, in tonnes.