29 0000 – 0000 Stuttgart, January 2005 Ga lley Pro of Verh. Internat. Verein. Limnol. Landscape-scale patterns of nitrogen fixation by cyanobacteria A. Patoine, M. Graham, P.R. Leavitt, and R. Hesslein Introduction Cyanobacteria blooms are characteristic of eutrophic lakes and are often a threat to water consumption (MILLIE et al. 1999). Lakes of the prairie ecozone of Canada are particularly prone to cyanobacteria blooms harbouring nitrogen-fixers (Aphanizomenon, Anabaena, Gloeotrichia spp.), in part because of phosphorus (P)-rich soils and high irradiance regimes (BARICA 1987). However, despite the prevalence of cyanobacteria in phytoplankton of some lakes, there is little agreement on whether nitrogen fixed from the atmosphere (N2) is quantitatively incorporated into the aquatic food web. On one hand, their colonial nature and potential toxicity make heterocystous cyanobacteria poor food for zooplankton (GHADOUANI et al. 2004). On the other hand, historical correlations between cyanobacterial abundance and stable isotope signals in plankton and lake sediments (TALBOT 2001) suggest that atmospheric N2 fixed by cyanobacteria may be an important source of nitrogen (N) to lake ecosystems. Little is known of how the importance of N sources may vary as a function of lake position within the greater landscape. Many chemical properties of lakes are strongly influenced by the position of a lake within the hydrologic or topographic landscape, such that headwater lakes tend to exhibit lower nutrient concentrations and algal densities than do downstream sites (SORANNO et al. 1999). Based on these patterns, we hypothesized that the intensity of N2 fixation by cyanobacteria should increase with distance from headwaters, and that fixed N2 should be more completely incorporated into the primary and secondary consumers of downstream lakes. To test these hypotheses, we analyzed seasonal variability in N isotope content of algae and zooplankton communities in five chained prairie lakes during 1995. Nitrogen fixation and assimilation by zooplankton should be reflected by concomitant declines in the 15N contents of algae and zooplankton, as light atmospheric N2 (δ15N = 0‰) is incorporated into relatively 15N-rich consumers (5–10‰). Umbruch Key words: landscape, cyanobacteria, N fixation, N concentration Material and methods Site description – The Qu’Appelle River drains 53,000 km2 of predominantly (76%) farmland as it flows eastward ~340 km from Lake Diefenbaker reservoir through a series of eight fluvio-glacial lakes, including Buffalo Pound, Katepwa and Crooked lakes (Fig. 1). The effective drainage area increases regularly with lake order (Fig. 2A). At the same time, soluble reactive phosphorus concentrations increase, while N:P ratios of dissolved nutrients decline as a function of lake position (Fig. 2B-C). Cyanobacteria are a common feature of Qu’Appelle Valley lakes (BARICA 1987) and were present before European settlement (HALL et al. 1999). Sampling and analyses – Diefenbaker, Buffalo Pound, Last Mountain, Katepwa and Crooked lakes were sampled bi-weekly May-August 1995. Integrated water samples from the surface to one meter above the sediments were used for phytoplankton taxonomic analyses and heterocyst counts. Water samples for N-isotopic analyses of particulate organic matter (δ15NPOM) were screened through 153-µm mesh net and filtered onto glass fiber filters. The screening procedure was effective at removing crustaceans and rotifers, while maintaining a representative algal assemblage characteristic of each lake (GRAHAM 1997). Zooplankton were collected with a 243-µm mesh conical plankton net. Adult zooplankton were sorted to species (Daphnia galeata mendotae, Leptodiaptomus sp., Diacyclops thomasi, Leptodora kindtii), dried, and frozen until isotope analyses. Mass spectrometry analyses of dried POM and zooplankton followed HESSLEIN et al. (1989). Isotope content (δ15N) was defined as permille (‰) deviation from the atmospheric standard and was calculated as: [(Rsample-Rstandard)/Rstandard] × 1000, where R is the isotope ratio (15N/14N). 0368-0770/05/0029-1756 $ 2.00 ©2005 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart Verh. Internat. Verein. Limnol. 29 Ga lley Pro of 2 Fig. 1. Qu’Appelle Valley drainage basin and lakes included in this study. Results The 15N content of algae (as POM) showed significant variations among lakes (Kruskal-Wallis Test Statistic = 9.9, P = 0.043), with lower values and greater seasonal variability in smaller, more productive lakes (Buffalo Pound, Katepwa, Crooked) than in larger Diefenbaker and Last Mountain Lakes (Fig. 2D). Most of this variability was inversely related to the absolute density of N2-fixing cyanobacteria (r2 = 0.76, P<0.001, Fig. 3A), suggesting that cyanobacteria influenced the overall isotopic signature of POM. Further, most of this trend was associated with small productive lakes that exhibited a wide range of cyanobacteria densities (103 to 105 cells mL–1) relative to larger sites (<104 cells mL–1). Fixed atmospheric N2 was assimilated into zooplankton only in downstream Katepwa and Crooked lakes (Fig. 4). At these sites, invertebrate d15N declined significantly as a function of the density of N2-fixing cyanobacteria. In contrast, there was no evidence of any negative relationship between N2-fixers and zooplank- ton isotopic signature in upstream Buffalo Pound Lake, suggesting zooplankton were not using cyanobacteria as a principal food source. Over the entire range of N2-fixing cyanobacteria relative density, the N-isotopic signature of zooplankton in upstream Buffalo Pound Lake tended to be lower relative to downstream lakes, probably because of inter-lake differences in baseline isotopic composition. No pattern was recorded in either of the large lakes (Diefenbaker and Last Mountain) due to very low cyanobacterial densities (Fig. 3A). Patterns of seasonal development of N2-fixing cyanobacteria differed substantially among lakes (Fig. 3B). In upstream Buffalo Pound Lake, blooms were smaller (<20 000 cells mL–1), and shorter-lived (<20 d) relative to downstream Katepwa and Crooked lakes. Estimates of the total amount of N2 fixed into each lake showed heterocystous cyanobacteria can contribute up to 8% of the average N pool (mean N concentration x lake volume) recorded in upstream Buffalo Pound lake, and more than 100% in downstream lakes (Table 1). Although these proportions were estimated Ga lley Pro of A. Patoine, Landscape – scale patterns 3 Fig. 2. Landscape patterns in A) drainage area, B) soluble reactive phosphorus (SRP), C) nitrogen to phosphorus mass ratio (as the sum of nitrate and ammonium relative to SRP) and D) nitrogen isotopic content of particulate organic matter (POM). Letters represent Diefenbaker Reservoir (D), Buffalo Pound Lake (B), Last Mountain Lake (L), Katepwa Lake (K), and Crooked Lake (C). Box and whiskers represent the 1995 seasonal variability. Centre vertical lines represent the median. Hinges represent the first and third quartiles and span the interquartile range (IQR). Upper/lower whiskers show the range of values that fall within the upper/lower inner fences i.e. upper/lower hinges +/– 1.5(IQR). Asterisks represent values between the inner fences and outer fences (upper/lower hinges +/– 3(IQR)). Empty circles represent values that fall outside the outer fences. from the growth season only and were derived from different methodologies (physiological estimates of N2 fixation rate, total N concentrations), they still suggest N2 fixation by heterocystous cyanobacteria is a more important N source in downstream lakes than in upstream lakes. Discussion If chemical properties of lake water are determined in part by a lake’s position in the land- scape, then so should some of its biological properties (SORANNO et al. 1999). Here, we showed that filamentous N2-fixing cyanobacteria, despite their apparently low palatability, were assimilated by zooplankton in the most downstream of a set of chained prairie lakes. This strong landscape pattern is consistent with observed declines in N:P ratios in downstream lakes (Fig. 2C), a factor known to favour development of heterocystous cyanobacterial blooms. In the Qu’Appelle lakes, this tendency to N limitation downstream arises because 4 Verh. Internat. Verein. Limnol. 29 Ga lley Pro of In contrast to lowland sites, atmospheric N2 did not appear to contribute substantially to food webs of upstream Buffalo Pound Lake, despite a high proportion of heterocystous cyanobacteria in phytoplankton (Fig. 4). This pattern likely arises because absolute densities of N2-fixers were low at all upstream sites (Fig. 3A, Table 1) and because inorganic N contents of these lakes were already relatively high (Fig. 2C). Thus, in this case, fixed N2 represented only a small proportion of total N standing stock (~8%) relative to downstream lakes and was less likely to influence the overall elemental composition of the consumer community. On-going analyses will reveal whether temporal patterns in N2 fixation are also observed in other years. In addition, mass-balance studies are being used to evaluate the relative importance of alternate N sources, including urban wastewater (δ15N ~17 ‰) and agricultural fertilizers (~0‰) (P.R. LEAVITT et al. unpubl. data). Conclusion Fig. 3. Nitrogen isotopic content of particulate organic matter as a function of density of nitrogenfixing cyanobacteria in Qu’Appelle Valley lakes (A) and seasonal development of nitrogen-fixing cyanobacteria (B). Numbers represent lake order on a West to East transect (1: Diefenbaker Reservoir; 2: Buffalo Pound Lake; 3: Last Mountain Lake; 4: Katepwa Lake; 5: Crooked Lake). Regression line follows the equation Y = 19.4 – 1.5 (log X), r2 = 0.76, P<0.001. lakes may act as net exporters of dissolved P (KENNEY 1990), leading to an “amplification” process whereby P accumulates in downstream lakes. The importance of N2 fixation in downstream lakes suggests internal sources of N such as sediment return were insufficient to alleviate N-limited phytoplankton growth (LEVINE & SCHINDLER 1992). Despite the filamentous and unpalatable nature of cyanobacteria, N from these phytoplankton is apparently incorporated into zooplankton communities. However, the degree of assimilation of fixed N2 by invertebrates was dependent on lake position in the landscape and was greatest in downstream lakes where low N:P ratios promoted dense and long-lived populations of N2-fixing cyanobacteria. At present, it is unclear whether fixed N2 is directly incorporated from cyanobacteria into invertebrates or whether it is first recycled from cyanobacteria into more palatable algae or bacteria. References BARICA, J., 1987: Water quality problems associated with high productivity of prairie lakes in Canada: a review. – Water Qual. Bull. 12: 107–115. GHADOUANI, A., PINEL-ALLOUL, B., PLATH, K., CODD, G.A. & LAMPERT, W., 2004: Effects of Microcystis aeruginosa and purified microcystin-LR on the feeding behavior of Daphnia pulicaria. – Limnol. Oceanogr. 49: 666–679. GRAHAM, M.D., 1997: Omnivory and selective feeding by zooplankton along a lake production gradient: complementary 15N isotope and gut pigment analyses. – M.Sc. Thesis, Univ. of Regina, Regina, Canada. HALL, R.I., LEAVITT, P.R., QUINLAN, R., DIXIT, A.S. & SMOL, J.P., 1999: Effects of agriculture, urbanization, and climate on water quality in the northern Great Plains. – Limnol. Oceanogr. 44: 739–756. Ga lley Pro of A. Patoine, Landscape – scale patterns 5 Fig. 4. Nitrogen isotopic content of Daphnia (bottom row), Copepoda (middle row) and Leptodora (top row) as a function of the relative density of cyanobacteria in upstream Buffalo Pound Lake (left colum) and downstream lakes (middle and right column). Numbers represent sampling events, from 1 (May 4) to 9 (August 31). Missing sampling events indicate that lake zooplankton densities were too low for sufficient material to be collected for mass-spectrometry analyses. Coefficients of determination and associated probability values are indicated. HESSLEIN, R.H., FOX, D.E. & CAPEL, M.J., 1989: Sulfur, carbon, and nitrogen isotopic composition of fish from the Mackenzie River delta region and other Arctic drainages. Canadian Data Report of Fisheries and Aquatic Sciences 728, iv + 11 pp. KENNEY, B.C., 1990: Dynamics of phosphorus in a chain of lakes: the Fishing Lakes. National Hydrology Institute, Paper no. 44, 18 pp., Saskatoon, Saskatchewan, Canada. LEVINE, S.N. & SCHINDLER, D.W., 1992: Modification of the N:P ratio in lakes by in situ processes. – Limnol. Oceanogr. 37: 917–935. MILLIE, D.F., DIONIGI, C.P., SCHOFIELD, O., KIRKPATRICK, G.J. & TESTER, P.A., 1999: The importance of understanding the molecular, cellular, and ecophysiological bases of harmful algal blooms. – J. of Phycol. 35: 1353–1355. SORANNO, P.A., WEBSTER, K.E., RIERA, J.L., KRATZ, T.K., BARON, J.S., BUKAVECKAS, P.A., KLING, G.W., WHITE, D.S., CAINE, N., LATHROP, R.C. & LEAVITT, P.R., 1999: Spatial variation among lakes within landscapes: ecological organization along lake chains. – Ecosystems 2: 395–410. TALBOT, M.R. 2001: Nitrogen Isotopes in Palaeolimnology. – In: LAST, W.M. & SMOL, J.P. (Eds.): Tracking Environmental Change Using Lake Sediments, Vol. 2, 6 Verh. Internat. Verein. Limnol. 29 Ga lley Pro of Table 1. Estimates of the amount of nitrogen fixed by cyanobacteria during the 1995 growing season, May-August. Computations are based on an average nitrogen fixation rate of 5.91 × 10–18 mol N2 s–1 heterocyst –1, and an average cell length of 7 µm (WALSBY 1985). Nitrogen fixation was assumed to occur uniformly during six hours per day. Buffalo Pound Katepwa Crooked Lake volume (millions of cubic meters) 87.5 233.2 120.9 Nitrogen-fixing cyanobacteria density (millions 5,132 33,757 44,993 per cubic meter of water, 1995 seasonal average) Number of heterocysts per 100 µm length of 2.1 6.9 7.1 filament (1995 seasonal average) Time span during which heterocysts were present 93 76 71 in collected samples (days) Estimated amount of nitrogen fixed by cyano22 1033 686 bacteria (tonnes) Estimated 1995 total nitrogen concentration 3,246 2,654 5,120 (mg m–3)* Estimated amount of nitrogen (tonnes) 284 619 619 Proportion of nitrogen atmospherically derived 8% 169% 111% * TN concentrations for 1995 were estimated from total inorganic nitrogen (TIN, nitrogen and ammonium) and average TIN/TN ratios of 0.14, 0.18 and 0.10 for Buffalo Pound, Katepwa, and Crooked Lakes respectively in subsequent years (1998, 1999, 2002). Physical and Geochemical Methods: 401–439. Kluwer Academic Press, The Netherlands. WALSBY, A.E., 1985: The permeability of heterocysts to the gases nitrogen and oxygen. – Proc. of the Royal Soc. of London B 226: 345–366. Authors’ addresses: A. PATOINE, M. GRAHAM, P.R. LEAVITT, Department of Biology, University of Regina, Regina, Saskatchewan, S4S 0A2, Canada. E-mail: [email protected] R. HESSLEIN, Fisheries and Oceans Canada, Freshwater Institute, Winnipeg, Manitoba, R3T 2N6, Canada.
© Copyright 2026 Paperzz