GLOBAL TEMP

    The rise in CO2 has a direct relationship on the rise in global temperature, which is expected to increase 2-4 degrees Celcius by 2100.("US EPA", 2012)  Fortunately the rise in CO2 not only increases temperature world wide, but it also increases the rate of photosynthesis, and therefore mangrove biomass.  With their special ability to fix carbon even more than terrestrial tropical forests, the increase in mangrove biomass will in return reduce carbon. (Alexander et al, 2011) 
     Mangrove trees are very temperature sensitive and do not generally grow in areas colder than 20 degrees Celcius. (Alongi, 2002) With an increase in global temperatures, which is expected to be even higher in higher latitudes, mangroves may be able to grow further up coasts, where the habitat has not been too degraded.  

 (The increase in CO2 and global temperature will increase mangrove abundance and in return decrease C02 and temperature.)

It is possible that an increase in droughts will impact the mangroves even more than a  temporary increase in photosynthesis, leading to a more negative outcome that includes even higher CO2 and higher rising temperatures. 

 The increase in global temperature will negatively impact mangrove abundance further increasing CO2 and temperature.

It does seem likely that global temperatures will increase photosynthesis at first, and therefore mangrove biomass, however photosynthesis has an optimal temperature, and tropical forests have already been found to be close to the maximum.(Billo pers comm) This implies that the scenario could be a temporary increase in mangroves and not a possibility to be relied on.

With an increase in coastal protection world-wide, it is possible that mangroves could expand  to higher latitudes, as they have already been seen to move inland, but there is no evidence of this yet.  There will also be a loss to mangrove areas that are too dry in the south, which could lead to an equilibrium or a loss overall, with the positive feedback loop in figure 2 dominating.