Nowhere is the human impact on this planet's chemistry more evident than in the industrial capture of atmospheric nitrogen for use in agricultural fertilizers. By increasing crop growth and reducing hunger, this Haber-Bosch process is said to be responsible for the current explosion of the human population [Smil, 1999]. Today, humans roughly double the input of fixed nitrogen to the global land surface, and rising human population will surely increase the global demand for fertilizer during the remainder of this century [Tilman et al., 2001]. Industrial fertilizer has released green plants from the nutrient that most limits their growth on land and in the oceans, a huge human perturbation of the planet's biogeochemistry [Schlesinger, 2004].
 Unfortunately, we have made few plans to control the fate of this unnatural supply of nitrogen in the environment. Indeed, we have only a limited understanding of where it all goes. Compared to the global production of fertilizer (100 TgN/yr), and other human sources of reactive N (56 TgN/yr [Galloway et al., 2004]), rivers appear to carry a burden of nitrogen that is only ∼20 TgN/yr greater than 100 years ago [Green et al., 2004; Bouwman et al., 2005]. The remaining input must remain in plants or soils, percolate to groundwater, or return to the atmosphere via a variety of processes, especially denitrification. Nearly all ecosystem budgets, at large and small scales, show this imbalance of nitrogen receipts and subsequent losses to surface waters [Van Breemen et al., 2002].
 Since nitrate is so soluble, hydrologists will be closely tied to all attempts to better understand the fate of fertilizer nitrogen in the environment. Increased river flux of various forms of nitrogen (ammonium, nitrate, dissolved organic N) is well documented, leading to excessive nitrogen concentrations in coastal waters and estuaries and to regional hypoxia in these waters [Caraco and Cole, 1999; Green et al., 2004]. A synoptic view of stream ecosystems links what a farmer does in a cornfield in Iowa to what a shrimp fisherman can no longer do in the Gulf of Mexico. Appropriate environmental policy and management should embrace catchments rather than political boundaries, and will require coordination between and across jurisdictions that have not traditionally talked to one another.
 Maintenance and restoration of small streams should become a high priority of land management, both rural and suburban. Already we have seen several reports of the importance of plant and microbial processes that remove nitrogen from first-order streams [e.g., Peterson et al., 2001; Bernhardt and Likens, 2002]. Attempts to model nitrogen transport and removal in larger rivers and through river networks suggest that nitrogen dynamics in small tributaries can strongly affect the amount and timing of nitrogen export to coastal waters [Alexander et al., 2000; Seitzinger et al., 2002]. Linking site-specific studies of stream and river nitrogen cycling with catchment-scale models of nitrogen export is critical for effective management of river networks, and for prioritization of protection and restoration efforts [Pringle, 2001].
 While we have long understood the basic processes in the nitrogen cycle, there is still considerable uncertainty in the magnitudes and rates of nitrogen cycling at the watershed and river basin scale [Reckhow et al., 2004]. Watershed management must begin with geospatial analysis of inputs and losses of nitrogen and other substances, so as to identify hydrologic flow paths, fluxes, stores, and residence times and the parts of ecosystems that are critical to nitrogen transformations, from dissolved to particulate and gaseous forms [Vidon and Hill, 2004; Creed and Band, 1998]. For example, confined animal feedlot operations (CAFOs) have expanded enormously in recent years, greatly increasing the amount of fixed and reduced nitrogen in selected river basins. Given the variability, within and between river basins, in factors that affect nitrogen transformations and transport (such as soil type, vegetation, soil moisture, and depth to groundwater), and the space/time scales of interest, we currently lack the scientific basis for effective protection of surface waters from these significant sources of nitrogen pollution. The hydrologic sciences have an important role to play in developing watershed-based planning to manage the nitrogen and other biogeochemical cycles of human interest (e.g., P and Si).
 The global sink for nitrogen in groundwater is unknown, but potentially very large. Nitrate leaching has contaminated the well water in many areas of the Midwestern U.S. and in other parts of the world. Much of this nitrate is potentially subject to denitrification, returning to the atmosphere as N2. Hydrologic processes control how much and what form of nitrogen is transferred from surface pathways to groundwater. Studies of the subsurface storage, transport, and transformations of nitrogen are a high priority for the hydrologic community.
 At least half of the nitrogen entering river systems appears to be lost to denitrification on its way to the sea [Galloway et al., 2004]. To the extent that denitrification produces nitrous oxide (N2O) in groundwater and stream sediment environments, it is not the panacea to all problems of excessive fertilizer nitrogen in the environment. Nitrous oxide is a powerful greenhouse gas that shows a rapidly increasing concentration in Earth's atmosphere. The ratio of N2O to N2 produced during denitrification is poorly constrained. A reasonable estimate may be close to 5% [Weier et al., 1993], but some streams show much lower values [Mulholland et al., 2004]. Matson et al.  show how the judicious seasonal use of nitrogen fertilizer can reduce the flux of N2O from agricultural fields to the atmosphere. Stream management can be informed by studies showing what combination of factors will lead to the greatest denitrification, with the majority of the nitrogen release as N2.
 Surely, humans will continue to use nitrogen fertilizer to grow food, with estimates ranging as high as 165 TgN/yr by 2050 [Galloway et al., 2004]. We can try to use fertilizer with greater efficiency and with techniques that minimize its loss to aqueous environments [Matson et al., 1998]. With the increasing cost of energy it will behoove us to obtain several cycles of nitrogen uptake by crops, before it is converted back to N2 by soil microbes or lost to runoff. Proper river basin management through biogeochemistry can improve the quality of surface waters for human use and for the preservation of natural ecosystems. We have come a long way from thinking about rivers simply as conduits to the sea. Increasing human impacts on global chemical cycles will demand that the hydrologic sciences put its best efforts toward the management of natural waters for a sustainable future.