Nearly all state of the art coupled climate models and ocean general circulation models produce a colder than observed cold tongue in the eastern equatorial Pacific which affects their ability to predict seasonal to interannual climate variability. It has recently been shown that poor representation of radiative trapping by near surface phytoplankton blooms may be largely responsible for this chronic problem. A coupled biological-ocean general circulation (BOGCM) has been employed to investigate the biological feedbacks on to ocean circulation, especially the SSTs which can directly affect the coupled climate system. The feedbacks are asymmetric with respect to ENSO events; much stronger (relatively weak) during cold (warm) ENSO events. The feedbacks in the Indian and Atlantic Oceans also show interannual variability and improve SST simulations significantly at seasonal to interannual time-scales. Just as the climate variability displays a decadal time-scale, the ecosystem and its feedbacks to the SST and air-sea interaction also have a distinct decadal component. A hybrid coupled ocean-atmosphere model (HCM) is used to test the feedbacks found in the BOGCM. While the distribution of radiation in the water column produces the same feedbacks in the HCM, the details are more intricate especially in terms of the relationship between equatorial zonal winds and off-equatorial SSTs and between the cold tongue SSTs and the stratus deck. The consequences of the marine ecosystem feedbacks are important not only to climate variability through SST feedbacks but also through CO2 exchanges with the atmosphere. The details of the bio-climate feedbacks along with the preliminary results on the air-sea exchange of CO2 in the tropics for 1948-2001 and its potential consequences for future climate variability will be discussed.