There is an urgent need to understand how saltmarsh will respond to the changing environmental conditions that result from climate change and other anthropogenic influences. Saltmarsh response to changing environmental conditions is difficult to predict at the ecosystem level as changes depend on the complex responses of species and communities. I investigated the responses of saltmarsh plants at individual, species and simplified community level to altered environmental conditions, including altering flooding regimes to simulate sea level rise using a newly developed Tidal Inundation Machine, and different nutrient conditions to simulate coastal eutrophication. I measured the expression of functional traits in order to relate plant responses to potential ecosystem functioning. I found that the variation of traits within a species was highly variable, irrespective of treatment and this served to dampen the observed effect of flooding at the community level. The effects of flooding were modified by the addition of nutrients, although this was very context-dependent, and flooding served to modify the intensity and direction of species interactions. I also found that different genotypes had different sensitivities to environmental conditions (flooding and nutrients), even differing in the direction of their response. This has real-world consequences as I found that genetic composition differed between saltmarshes, with variation partially explained by flooding frequency. However, contrary to expectations, restored marshes did not differ from natural sites in their genetic diversity, even two years after restoration. These experiments were facilitated by the development of the Tidal Inundation Machine that was able to reproduce a true tidal cycle, as well as controlling for nutrient 4 concentrations, enabling me to study the combined effects of increased tidal inundation and nutrient enrichment. Overall, I found substantial variation in the responses of individual plants to changes in the environment. Sources of variation included neighbourhood composition, intra-specific trait variability and genotype. Collectively these represent a hierarchy of predictability of responses. This complexity will impact on our ability to predict responses to future change and highlights the need to better understand plants at the individual level before we can predict the response across entire ecosystems.