The globally significant green turtle (Chelonia mydas) population in the northern Great Barrier Reef is threatened by anthropogenic pressures, including climate change, habitat degradation, and indigenous harvest. Evidence suggesting the population is producing an extreme proportion of females due to increasing temperatures, coupled with temperature-dependent sex determination, is concerning. In response, and to explore management options, we developed two density-independent, stochastic stage- structured metapopulation models: a "Moderate Climate Model" and an "Extreme Climate Model". The models differ based on climate change projections by incorporating increased female hatchling sex ratios due to global warming and loss of nesting habitat due to sea level rise. The models were based on demographic data from field studies at major rookeries and regional foraging grounds and allowed for variation in operational sex ratios, management actions, and levels of indigenous harvest. Under the Moderate Climate Model, population size increased but could be vulnerable to overharvest of adult females. If overharvest was indicated, the harvest of a proportion of subadults rather than only adult females reduced population declines. Under the Extreme Climate Model, there was a steep population decline even without any harvest and harvesting subadults accelerated population decline due to the inclusion of subadult males. In the Extreme Climate Model, reversal of population decline depended on male turtles mating with an increased number of females, or management actions to substantially increase the number of male hatchlings produced.