The saved ecology animation classifies horizontal position using thirty five percent and sixty eight percent of canvas width. On a hypothetical one-thousand-pixel-wide canvas, the reef region spans three hundred fifty pixels, the middle region three hundred thirty and the lagoon region three hundred twenty. At half scale the displayed widths are one hundred seventy five, one hundred sixty five and one hundred sixty. These are pixel regions, not mapped Adriatic habitat areas or field observations. Random spawning and movement produce illustrative agents. The simulation does not establish that predator introduction, dredging or trapping would be safe or effective in a real ecosystem. A dredge click beyond forty percent of canvas width reduces the silt variable by four, with a minimum of four. Starting from eighteen, successive eligible clicks give fourteen, ten, six and four, then remain four. At ten pixels per silt unit, eighteen, ten and four produce bars one hundred eighty, one hundred and forty. This tool directly edits a stored variable and clears nearby burrow flags inside an eighty-pixel radius. Its labels are software assumptions, not calibrated sediment depths or an environmental recommendation. The predation-efficiency label divides cumulative kills by kills plus currently living crabs, then rounds a percentage. A hypothetical twenty kills and eighty living gives twenty percent. Twenty kills and twenty living gives fifty percent. At five pixels per percentage point those bars measure one hundred and two hundred fifty. The denominator is not all crabs ever spawned, despite a guard checking that spawning count is positive. Monthly time advances by zero point zero two five per update call, so it depends on frame cadence rather than elapsed real seconds. These counters and proxy labels are not validated estimates of ecological control outcomes.