Moonem Timbes

Moonem Timbes puts you in charge of a deteriorating river valley. You have one decision per year, a fixed number of years, and a target that is specific enough to be meaningful and difficult enough to be real. Every action you take lands years after you take it. Release the wolf pack this year: the herd shifts its movement in year two, mesopredator suppression follows in year three. Plant willow on the bank and those stems need four more years and sustained low browse pressure before they clear browse height and start holding the soil. The valley you are reading right now reflects decisions made years ago, and curves that still look like they are falling may already have been reversed. Moonem Timbes is built around that lag, the one thing every conservationist knows and almost nothing has ever made you feel in real time. Under the surface runs a system of fourteen coupled equations modelling predators, prey, mesopredators, beaver colonies, woody stems, escaped willow, bank stability, songbirds, and scavengers. A live food-web screen shows every relationship weighted by real computed values this year. A Tracer screen walks backward through the model's causal chain, up to four steps deep, to tell you exactly which decision in which year produced what you are seeing now. All game state is stored entirely on-device. No accounts, no sync, no analytics. Six curated scenarios each hand you a broken valley and a specific target: a system without a wolf pack for thirty years, a pack that worked too well and left the herd crashing, a dry decade where holding the line is the whole job, a beaver question that depends on getting willow tall enough first. A free sandbox lets you configure every starting condition, enter a seed to share your exact valley with someone else, and run it with no target and no year limit.