Read the explanation
This is a two-pool source toy model, not evolution, genetics or a planetary forecast. At defaults biomass twenty, nutrient pool eighty, growth rate point zero eight, half-saturation thirty and efficiency point five, growth is the smaller of forty and point zero eight times twenty times eighty divided by one hundred ten: about one point one six three six four. Mortality point zero four times twenty is point eight. Bars at one hundred pixels per source flow unit show about one hundred sixteen point three six and eighty pixels. The biomass update is twenty plus growth minus mortality, about twenty point three six three six four. These are source-defined units, not calibrated ecology. The same first update removes growth divided by efficiency, about two point three two seven two seven, from the nutrient pool. It returns mortality times recovery, point eight times point eight, or point six four. At one hundred pixels per source nutrient-flow unit, bars show about two hundred thirty-two point seven three and sixty-four pixels. Nutrients therefore change from eighty to about seventy-eight point three one two seven three. The exported turn accumulator adds raw mortality, point eight, not the point six four actually returned; the source label recycled should not be read as that nutrient-return amount. No physical conservation or ecological calibration is established. The source records the initial state and each of one hundred twenty discrete updates, giving one hundred twenty-one rows per run. Bars compare one hundred twenty updates with one hundred twenty-one recorded states at four pixels per count. The baseline sets mortality to zero; turnover uses the chosen mortality. With zero chosen mortality the two runs coincide and accumulated mortality is zero. The chart uses a fixed biomass domain zero to eighty. Stored parameters are not restored: reload starts the defaults and overwrites storage. Export records all rows and limitations; no species, external inputs, genetics or evolution are simulated.