Wetlands are often highly effective nitrogen (N) sinks. In the Lake Waco Wetland (LWW), near Waco, Texas, USA, nitrate (NO 3 -) concentrations are reduced by more than 90% in the first 500 m downstream of the inflow, creating a distinct gradient in NO 3 -concentration along the flow path of water. The relative importance of sediment denitrification (DNF), dissimilatory NO 3 -reduction to ammonium (DNRA), and N 2 fixation were examined along the NO 3 -concentration gradient in the LWW. ''Potential DNF'' (hereafter potDNF) was observed in all months and ranged from 54 to 278 lmol N m -2 h -1 . ''Potential DNRA'' (hereafter potDNRA) was observed only in summer months and ranged from 1.3 to 33 lmol N m -2 h -1 . Net N 2 flux ranged from 184 (net denitrification) to -270 (net N 2 fixation) lmol N m -2 h -1 . Nitrogen fixation was variable, ranging from 0 to 426 lmol N m -2 h -1 , but high rates ranked among the highest reported for aquatic sediments. On average, summer potDNRA comprised only 5% (±2% SE) of total NO 3 -loss through dissimilatory pathways, but was as high as 36% at one site where potDNF was consistently low. Potential DNRA was higher in sediments with higher sediment oxygen demand (r 2 = 0.84), and was related to NO 3 -concentration in overlying water in one summer (r 2 = 0.81). Sediments were a NO 3 -sink and accounted for 50% of wetland NO 3 -removal (r 2 = 0.90). Sediments were an NH 4 + source, but the wetland was often a net NH 4 + sink. Although DNRA rates in freshwater wetlands may rival those observed in estuarine systems, the importance of DNRA in freshwater sediments appears to be minor relative to DNF. Furthermore, sediment N 2 fixation can be extremely high when NO 3 -in overlying water is consistently low. The data suggest that newly fixed N can support sustained N transformation processes such as DNF and DNRA when surface water inorganic N supply rates are low.

Data from five sampling stations for 23 dates in 1 yr show that annual phytoplankton production in Lake Chapala is low (80 g C m-*) and governed by high inorganic turbidity. In the most turbid region of the lake, Secchi transparency averaged 0.2 m and the vertical light (PAR) attenuation coefficient (7") averaged 9.7 m-l; in the least turbid region, Secchi transparency averaged 0.7 m (q" averaged 2.3 m-l).Phytoplankton at the shallowest and most turbid station were most productive per unit volume, while the least turbid station, with a deep circulating water column, had the lowest volume-based production. There was a considerable rainy vs. dry season difference in water transparency and in production. The lake became less turbid during the rainy season due to increased depth and lessened sediment resuspension. The annual lakewide production C.V. was 34% with the greatest day-today variation during the rainy season.Phytoplankton Chl a averaged 5.4 mg m-3 of the mixed water column for all regions. Chl a at the shallowest and most turbid station averaged almost twice that of the other stations. Chl a increased through the rainy season (as transparency and inorganic N content increased). Variation in inorganic N accounted for the greatest variation in Chl a at each region.Water, pigments, dissolved organic matter, and particles attenuate sunlight passing the surface of lakes and reservoirs. Light attenuation by phytoplankton pigments has received much attention, because of the selflimitation of photosynthesis imposed by the photosynthesizing organisms (Talling 1957, 197 1;Bannister 1974a). Neglected has been the role of sunlight attenuation through scattering and absorption by clay and silt suspensoids. Yet, for many reservoirs and Acknowledgments

We investigated whether rates of net primary production (NPP) and biomass turnover of floating grasses in a central Amazon floodplain lake (Lake Calado) are consistent with published evidence that CO 2 emissions from Amazon rivers and floodplains are largely supplied by carbon from C4 plants. Ground-based measurements of species composition, plant growth rates, plant densities, and areal biomass were combined with low altitude videography to estimate community NPP and compare expected versus observed biomass at monthly intervals during the aquatic growth phase (January-August). Principal species at the site were Oryza perennis (a C3 grass), Echinochloa polystachya, and Paspalum repens (both C4 grasses). Monthly mean daily NPP of the mixed species community varied from 50 to 96 g dry mass m À2 day À1 , with a seasonal average (AE1SD) of 64 AE 12 g dry mass m À2 day À1 . Mean daily NPP (AE1SE) for P. repens and E. polystachya was 77 AE 3 and 34 AE 2 g dry mass m À2 day À1 , respectively. Monthly loss rates of combined above-and below-water biomass ranged from 31% to 75%, and averaged 49%. Organic carbon losses from aquatic grasses ranged from 30 to 34 g C m À2 day À1 from February to August. A regional extrapolation indicated that respiration of this carbon potentially accounts for about half (46%) of annual CO 2 emissions from surface waters in the central Amazon, or about 44% of gaseous carbon emissions, if methane flux is included.

No abstract