Dams, floodplain land use, and riparian forest conservation in the semiarid Upper Colorado River Basin, USA.

Land and water resource development can independently eliminate riparian plant communities, including Fremont cottonwood forest (CF), a major contributor to ecosystem structure and functioning in semiarid portions of the American Southwest. We tested whether floodplain development was linked to river regulation in the Upper Colorado River Basin (UCRB) by relating the extent of five developed land-cover categories as well as CF and other natural vegetation to catchment reservoir capacity, changes in total annual and annual peak discharge, and overall level of mainstem hydrologic alteration (small, moderate, or large) in 26 fourth-order subbasins. We also asked whether CF appeared to be in jeopardy at a regional level. We classified 51% of the 57,000 ha of alluvial floodplain examined along >2600 km of mainstem rivers as CF and 36% as developed. The proportion developed was unrelated to the level of mainstem hydrologic alteration. The proportion classified as CF was also independent of the level of hydrologic alteration, a result we attribute to confounding effects from development, the presence of time lags, and contrasting effects from flow alteration in different subbasins. Most CF (68% by area) had a sparse canopy (50% canopy cover occupied <1% of the floodplain in 15 subbasins. We suggest that CF extent in the UCRB will decline markedly in the future, when the old trees on floodplains now disconnected from the river die and large areas change from CF to non-CF categories. Attention at a basinwide scale to the multiple factors affecting cottonwood patch dynamics is needed to assure conservation of these riparian forests.

Patterns of nitrogen accumulation and cycling in riparian floodplain ecosystems along the Green and Yampa rivers.

Patterns of nitrogen (N) accumulation and turnover in riparian systems in semi-arid regions are poorly understood, particularly in those ecosystems that lack substantial inputs from nitrogen fixing vegetation. We investigated sources and fluxes of N in chronosequences of riparian forests along the regulated Green River and the free-flowing Yampa River in semi-arid northwestern Colorado. Both rivers lack significant inputs from N-fixing vegetation. Total soil nitrogen increased through time along both rivers, at a rate of about 7.8 g N m(-2) year(-1) for years 10-70, and 2.7 g N m(-2)year(-1) from years 70-170. We found that the concentration of N in freshly deposited sediments could account for most of the soil N that accumulated in these floodplain soils. Available N (measured by ion exchange resin bags) increased with age along both rivers, more than doubling in 150 years. In contrast to the similar levels of total soil N along these rivers, N turnover rates, annual N mineralization, net nitrification rates, resin-N, and foliar N were all 2-4 times higher along the Green River than the Yampa River. N mineralization and net nitrification rates generally increased through time to steady or slightly declining rates along the Yampa River. Along the Green River, rates of mineralization and nitrification were highest in the youngest age class. The high levels of available N and N turnover in young sites are not characteristic of riparian chronosequences and could be related to changes in hydrology or plant community composition associated with the regulation of the Green River.