Challenges in advancing the study of the long-term wetland vegetation dynamics lie in the ability to capture the long-term features of wetland fluctuating environment, which impose various stresses to, while simultaneously being modified by, the vegetation. This dissertation describes two modeling examples, which collectively pioneer: (1) the use of precipitation-driven stochastic water levels and soil moisture as the key variables controlling wetland plant growth, directly through water availability and indirectly through the dynamics of oxygen and nitrogen; and (2) the incorporation of feedbacks between wetland plants and their environments through transpiration loss, long distance oxygenation, and litter dynamics. The first model considers competition outcomes between plant species differently trading-off flood resistance with growth in a dry type of wetlands having highly fluctuating soil saturation with minimal standing water, where hydrologic and vegetation dynamics are coupled via transpiration and ecosystem carrying capacity and the effects of light, oxygen, and nitrogen are implicitly considered. The feedbacks result in long memory in the water level signals, i.e., persistence in floods and droughts, and consequently long-range self-similar characters in vegetation dynamics. The statistical structures of the modeled water table agree well with empirical data, exhibiting the power-law character. The second model considers dynamics of emergent herbaceous monoculture in a wetter type of wetlands having intermittent to almost perennial standing water, which allows for decoupling water level fluctuations from vegetation dynamics. Effects of multiple stresses including drought, flood, shading and nitrogen limitation as well as plant feedbacks onto the environment are considered through explicit accounting of carbon (C) and nitrogen (N) distribution between pools of plants, soil, and microbes under a closed N cycle assumption. Vegetation growth is optimized through dynamic allocation of C and N gain. The model can capture changes in sawgrass traits, structure, and productivity under various hydrologic and nutrient regimes consistent with field observation in the southern Everglades and provide insights into physiological features that enable such plant adaptations, suggesting its potential toward the quantification of the long-term wetland vegetation dynamics.