Historically, aquatic ecosystems have been thought to be relatively isolated, and energetically self sufficient. That is, researchers assumed that the aquatic food web and its metabolism was supplied, ultimately, by primary production from the phytoplankton and aquatic plants in the water body itself. A number of lines of evidence demonstrate that aquatic ecosystems are significantly subsidized by inputs of terrestrial organic matter. The evidence includes the persistent supersaturation of carbon dioxide (or undersaturation of dissolved oxygen) in surface waters (Cole et al., 1994, Science); the observation that measurements of respiration exceed that of primary production (del Giorgio et al., 1997, Nature) and the mass balance of organic C in lakes and rivers (Caraco and Cole, 2004). Cole is investigating this connection between the terrestrial and aquatic C cycles and its consequences in several projects.

Figure Legend: Rivers (upper panel) and Lakes (lower panel) tend to be supersaturated in CO₂. That is, in the absence of additional inputs of CO₂ surface waters would be at equilibrium with the overlying atmosphere. The data are plotted as "Relative Saturation". At equilibrium Relative Saturation =1.0. A value of 4 means the surface water has 4 time more CO₂ than can be explained by equilibrium with the atmosphere. A value of -4 means the water is undersaturated, and has 1/4th the equilibrium value. The data for lakes includes 1835 lakes with a global distribution; the data for rivers includes 80 rivers, also with a global distribution. The frequency is given as number of samples.