In the early 1990s researchers at the Arctic LTER discovered the answer to a long-standing question of why the recently measured and modeled rates of net ecosystem productivity on land were so much higher than the rates of carbon accumulation integrated over time as found in soil and peat cores. The researchers discovered that carbon dioxide and methane dissolved in groundwater in very high concentrations were transported to streams and rivers and released to the atmosphere (Kling et al. 1991). This lateral transport of carbon away from land and into surface waters was missing from the calculation of net ecosystem productivity, and explained why short-term measurements of net productivity were higher than the long-term, integrated rates of carbon storage: the carbon was taken out of the atmosphere by plants through photosynthesis, but when the plants died only part of their carbon was stored in soil (less than was originally produced). The rest was converted by microbes to carbon dioxide and methane, some of which was transported into lakes and streams. Thus, lakes and rivers are significant conduits for the delivery of carbon dioxide and methane originating on land to the atmosphere, accounting for at least 20% of the net carbon balance for the entire Arctic (Kling et al. 1991).

The greater contribution to scientific understanding, however, was the realization that this gas conduit operates worldwide but is often overlooked as an important source of greenhouse gases to the global atmosphere. The first paper describing this global application (Cole et al. 1994) has now been cited well over 300 times, and has spawned a huge amount of research on the linkages of terrestrial and aquatic habitats with respect to the carbon budgets of these ecosystems. A final outgrowth of this research has been its impetus to consider land-water interactions and, especially, the interaction of lakes and streams to understand the functioning of an entire landscape (Kling et al. 2000). Both of these scientific advances -- carbon cycling between land and surface waters and the integration of land, lakes, and streams to understand larger-scale landscape function -- now routinely form the basis of special sessions at scientific meetings, workshops, and symposia.