1. The sources of nitrogen for phytoplankton were determined for a bloom-prone lake as a means of assessing the hypothesis that cyanobacteria dominate in eutrophic lakes because of their ability to fix nitrogen when the nitrogen : phosphorous (N : P) supply ratio is low and nitrogen a limiting resource. 2. Nitrogen fixation rates, estimated through acetylene reduction with 15 N calibration, were compared with 15 N-tracer estimates of ammonium and nitrate uptake monthly during the ice-free season of 1999. In addition, the natural N stable isotope composition of phytoplankton, nitrate and ammonium were measured biweekly and the contribution of N 2 to the phytoplankton signature estimated with a mixing model. 3. Although cyanobacteria made up 81-98% of phytoplankton biomass during summer and autumn, both assays suggested minimal N acquisition through fixation (<9% for the in-situ incubations; <2% for stable isotope analysis). Phytoplankton acquired N primarily as ammonium (82-98%), and secondarily as nitrate (15-18% in spring and autumn, but <5% in summer). Heterocyst densities of <3 per 100 fixer cells confirmed low reliance on fixation. 4. The lake showed symptoms of both light and nitrogen limitation. Cyanobacteria may have dominated by monopolizing benthic sources of ammonium, or by forming surface scums that shaded other algae.

No abstract

To test the hypothesis that N:P supply ratios influence phytoplankton species composition, and particularly that cyanobacteria are favored by a low ratio, mesocosms at one pelagic and two littoral sites within the Experimental Lakes Area, Ontario, were fertilized for 10-12 weeks with similar amounts of P but different amounts of N. Total N:P supply (LN:LP) ratios (fertilizer plus natural inputs) ranged from 8:1 to 50:1. Nitrogen deficiency was detected in all mesocosms with LN:LP ratios <17:1, but N2-fixing Anabaena gained dominance only in the low-N:P pelagic mesocosms, and only in late summer. Cryptophytes and (or) chlorophytes dominated littoral mesocosms at all N:P ratios, while Pseudoanabaena catenata, a nonheterocystous cyanobacterium, was the late-summer dominant in pelagic mesocosms with LN:LP ratios >17:1. Canonical correlation analysis related cyanobacterial dominance to high P and low CO2 availability. Low light intensities and low N:P ratios also favored heterocystous (but not nonheterocystous) cyanobacteria. Total phytoplankton biomass and productivity increased with LN:LP ratio, while periphyton growth was maximal at low N:P ratios. Nitrogen limitation of phytoplankton may encourage blooms of N2 fixers or drive productivity down to the sediment surface where N is more available.

Phosphoric acid labelled with 32P was added to the epilimnion of experimentally eutrophied Lake 227, northwestern Ontario, in August 1978 to trace the lake's phosphorus cycle during late stratification and fall overturn. Radiophosphate was incorporated into bacteria and microphytoplankton (<10 μm in diameter) within minutes of introduction. Thereafter, the bulk of 32P exchange was between the microplankton, which typically held >90% of 32P, dissolved phosphate, and a mobile subcompartment of "dissolved organic" phosphorus. Phytoplankton >10 μm in diameter and zooplankton acquired 32P label very slowly and contained <5% of the radiotracer. In addition, about half of dissolved organic phosphorus failed to acquire a 32P label over 17 d of incubation, indicating that this phosphorus may be functionally inert, at least at the time scale of biological phosphorus exchange. Movement of 32P to the littoral zone of Lake 227 was much slower than in lakes with well-developed macrophyte communities, but losses to deep sediments were similar, about 2% of 32P per day. Particles settled through the hypolimnion at a modal velocity of 28 cm∙d−1, losing 10% or less of their 32P during descent. Most of the released phosphorus was reincorporated into particles and sedimented. Less than 1% of the 32P that entered the hypolimnion in particles between August and October remained in the hypolimnion in late October. 32P release from Lake 227′s hypolimnetic sediments during late summer and fall of 1978 was <5% of the 32P that sedimented.

Water samples from six depths in the upper water column of Lake Valencia, Venezuela, were incubated in situ with acetylene at several depths corresponding to different light exposures. On five dates there was sufficient nitrogen fixation to define the light-response curve for samples taken at 2 and 100 cm. The light response of nitrogen fixation was modeled successfully with an equation previously developed for the light response of photosynthesis. All parameter values of the response curves for samples originating at a given depth were relatively stable across dates, but differed between depths. Rates of increase of N fixation with irradiance in the subsaturation range (a) were much lower than is typical of photosynthesis. While the general shape of nitrogen fixation curves was similar to that of photosynthesis curves, critical parameters were more sensitive to the light history (depth of origin) of cells than would be expected for photosynthesis.

1. Wind-induced sediment resuspension in shallow lakes affects many physical and biological processes, including food gathering by zooplankton. The effects of suspended sediment on clearance rate were determined for a dominant cladoceran, Daphnia carinata, and calanoid copepod, Boeckella hamata, in Lake Waihola, New Zealand. 2. Animals were incubated at multiple densities for 4 days in lake water containing different amounts of suspended lake sediment. Rates of harvest of major food organisms were determined for each sediment level (turbidity) from changes in net growth rate with grazer density. 3. Daphnia cleared all food organisms 7-40 lm in length at similar rates, but was less efficient in its removal of free bacteria, phytoplankton <7 lm, and large cyanobacterial filaments. Elevation of sediment turbidity from 2 to 10 nephelometric turbidity units (NTU) (63 mg DW L )1 added sediment) reduced Daphnia clearance of phytoplankton, heterotrophic flagellates and ciliates by 72-100%, and of amoebae and attached bacteria by 21-44%. Further inhibition occurred at higher turbidity. 4. Boeckella hamata removed microzooplankton primarily, rather than phytoplankton. The rate at which it cleared rotifers was reduced by 56% when turbidity was increased from 2.5 to 100 NTU. 5. In the absence of macrozooplankton, algal growth increased with sediment turbidity, suggesting that sediment also inhibits rotifer grazing. 6. As mid-day turbidity in Lake Waihola is ‡10 NTU about 40% of the time, sediment resuspension may play a major role in moderating energy flow and structuring pelagic communities in this lake.

In situ mesocosms in two Canadian Shield lakes were used to evaluate the contributions of inlake vs. external sources of nitrogen and phosphorus to nutrient budgets and N : P ratios. These mesocosms were designed to have variable exchange with sediments. Half were fertilized with N and P at a ratio great enough to ensure P limitation for most phytoplankton (atomic ratio, 33 : 1); the other half were fertilized at a ratio low enough to cause N limitation (4.4 : 1) in the absence of compensation mechanisms.For littoral mesocosms, sediments were a major source of N, but not of P. A comparison of mesocosms having sediments with one having a plastic floor indicated that sediment N return was derived largely from decomposing material at the sediment surface, rather than from deep sediments. Disproportionate returns of N from sediments, along with lower denitrification, reduced N limitation in the low N : P mesocosms. In pelagic mesocosms, which lacked sediment contact, N, fixation and thermocline entrainment late in the experiment were the principal internal N sources at low N : P.Biogeochemical mechanisms for reducing N : P in the water column at high N : P supply ratios were less effective than those that ameliorated N shortages at low N : P. The most important mechanism for reducing N : P was denitrification, while both N, fixation and sediment return raised N : P. We conclude that biogeochemical mechanisms allow N shortages to be more readily overcome than P shortages in Canadian Shield lakes. Due to the importance of sediments as an N source, it is important to scale mesocosms so that they have sediment: water ratios similar to those of the lakes they are designed to simulate.

Little is known about the recovery of lakes from eutrophication, especially as it effects the cycling of nutrients other than phosphorus. We fertilized a naturally oligotrophic lake (Lake 303) in the Experimental Lakes Area, northwestern Ontario, with nitrogen and phosphorus over two summers and examined the subsequent recovery using mass balance and large in situ mesocosms. While large amounts of ammonium and smaller amounts of dissolved inorganic carbon (DIC) were released from the lake's aerobic sediments during the first year of recovery, phosphorus input from the sediments was not detectable during the same period. Chlorophyll a concentration closely tracked total phosphorus concentration, and both returned to prefertilization levels within 1 yr. In contrast, ammonium, nitrate, DIC, and dissolved organic carbon (DOC) needed almost 2 yr to reach baseline levels, and dissolved organic nitrogen (DON) even longer. The results show that different elements may recover at different rates and that the release of stored pools of nitrogen and carbon in the sediments plays a major role in delaying the recovery of these elements.