**corresponding author: Phone: (615) 898-2660,
Email: kwalt@mtsu.edu.
Abstract: The effects of eutrophication on the formation and characteristics of limnetic organic aggregates (lake snow) were examined in a February 1997 experiment. Water samples were collected from a hard-water Middle Tennessee lake, returned to the lab, and placed in 1.2 l rotating cylinders. Varying amounts of a commercially available fertilizer were added to cylinders to create three known nutrient concentration treatments (0X = ambient, 1X = ambient + 10 mg l-1 total nitrogen, 5X = ambient + 50 mg l-1 total nitrogen). The effects of nutrient loading on the density and individual volumes of aggregates formed were determined after 48 h. Nutrient additions significantly increased both the numbers and sizes of aggregates formed. Aggregate densities increased by 78% and aggregate sizes increased by greater than seven fold between 0X and 5X treatments. The effects of nutrient additions on the numbers and sizes of organic aggregates formed in our experiment suggests that eutrophication will have a profound affect on water column dynamics and the vertical transport of organic and inorganic material in limnetic systems.
Introduction
Macroscopic organic aggregates (snow) composed of detritus, mineral particles, inorganic material, and organisms such as bacteria, algae, and protists are common in freshwater (e.g.; Grossart & Simon 1993) and marine environments (e.g.; Alldredge & Silver 1988). Snow typically forms as microscopic particles collide through differential settling or physical shear and adhere by the actions of various organic compounds (Alldredge & Silver 1988). Microbes frequently are orders of magnitude greater in aggregates than in surrounding waters (Alldredge & Youngbluth 1985, Lochte & Turley 1988, Simon et al. 1990). The settling of aggregates, such as organics, nutrients, and many dissolved molecules (e.g. heavy metals, pollutants) that adsorb to aggregate surfaces, can facilitate the transport of materials out of the euphotic zone (Alldredge & Silver 1988, Shanks & Edmonson 1990).
Lacustrine and marine snow is similar but can differ in regard to the formation and characteristics of aggregates. Typically, wind-induced turbulence is required for the formation of lake snow (Grossart & Simon 1993), but Brownian diffusion and differential settling also contribute to aggregate formation (Weilenmann et al. 1989). Solution chemistry strongly influences adhesion of microscopic particles to form snow in freshwater systems (Weilenmann et al. 1989). Increased calcium ion concentrations as found in hard-water lakes will act as a destabilizing agent and promote the natural coagulation of particles (Weilenmann et al. 1989). Also, the size (lake snow typically being smaller than marine snow) and microbial composition of aggregates can differ between freshwater and marine aggregates (Grossart & Simon 1993).
Nutrient loading, predominantly from non-point source pollution, and resultant eutrophication is a major problem in many freshwater systems. The ability of aggregates to accumulate and remove nutrients and other pollutants from the water column (Alldredge & Silver 1988, Weilenmann et al. 1989, Grossart & Simon 1993) suggest that the dynamics of snow particles would be affected by nutrient loading. In this paper we examine the physical effects nutrient loading has on the formation of organic aggregates in a limnetic environment.
Materials and Methods
Following the methods of Shanks & Edmondson (1989), lake snow was generated within rotating cylinders (1.2 l total volume) in the laboratory. Although cylinders do not mimic natural hydrodynamic regimes (Jackson 1994), cylinder-formed aggregates are similar to field aggregates in density, size, and composition (Shanks & Edmondson 1989). Lake water for experiments was collected on 4 February 1997 from Jefferson Springs on J. Percy Priest Lake just north of Murfreesboro, Tennessee, USA (86o 30' W, 36o 00' N). Water temperature at the collection site was 4.5 C. Nitrate, phosphate and total phosphorus levels determined using a hot kit were 0.8, 0.59, and 0.19 mg l-1, respectively.
In the lab, field-collected lake water was placed with a minimum of disruption into cylinders. Nutrient treatment levels representing 0X, no addition, 1X, 10 mg l-1 total nitrogen addition, and 5X, 50 mg l-1 total nitrogen addition, were established within cylinders by adding predetermined amounts of a commercial (A. Y. Schultz Mfg.) 20-30-20 plant food (total nitrogen equaling 16.4% urea, 0.4% nitrate and 3.2% ammonia) with micronutrients (0.1% iron, 0.05% manganese, 0.05% zinc). The range of treatment levels were selected to include total nitrogen amounts considered by the US Environmental Protection Agency to represent moderate to severe eutrophication. Cylinders containing the different nutrient treatments were allowed to rotate a total of 48 h on a 12/12 light/dark cycle. Aggregates typically formed within 30 minutes after the onset of rotation. The contents of each cylinder were photographed at the end of the 48 hour period to determine aggregate numbers and size. Photographs were taken using a Cannon AE1 and Kodak TMX 135 black and white film at an f-stop that permitted sufficient depth of field to visualize all aggregates within the cylinder. A black background behind the cylinders and 500 w floods on either side increased the resolution of individual aggregates.
Aggregate numbers and linear dimensions were determined from the 1:1 photographs of cylinders. All visible aggregates within 4 selected 10 pie sections randomly chosen from within each 90 quadrat in the photographs were enumerated. Using an ocular micrometer to determine aggregate size, the length and width of a haphazardly selected subset of aggregates (25 w/ each cylinder section) were measured under a dissecting microscope at 6X. If length and width were equal, aggregate volumes were calculated based on formulas for a sphere; if length exceeded width, aggregate volumes were calculated based on formulas for a right cylinder.
Results were analyzed by ANOVA and Ryan's Q multiple comparison tests (Day & Quinn 1985). Data were log transformed to satisfy homogeneity of variance assumptions. A nested design with counts from individual fields treated as subsamples was used to analyze aggregate density and size. All tests were run on an HP9000 using SPSS 6.1 (Norusis 1993) or SAS 6.04 and the GLM statistical routine (Joyner 1985).
Results
Numbers and sizes of aggregates formed over 48 h of rotation varied appreciably among nutrient treatment levels. The trend was for greater numbers of aggregates per liter and greater aggregate size in treatments with increased nutrient levels (Figures 1 & 2). Aggregate numbers were significantly different (F = 14.94; df = 2, 9; p < 0.005) between nutrient addition levels (Figure 1) and between randomly-chosen sections within cylinders (F = 4.86; df = 9, 36; p < 0.001). Densities of aggregates were significantly increased by over 78% in the 50 mg/l treatment (Ryan's Q, p < 0.05). Aggregate size also was significantly different (F = 48.43; df = 2, 9; p < 0.0005) between nutrient treatments (Figure 2) but not significantly different between randomly chosen cylinder sections (F = 1.85; df = 9, 288; p > 0.05). Sizes of individual aggregates significantly increased 5-fold between 0X and 1X treatments and 2.6-fold between 1X and 5X treatments (Ryan's Q, p < 0.05).
Discussion
The addition of nutrients in the form of a commercially
available fertilizer to water collected from J. Percy Priest Lake, TN resulted
in an increase in the density and size of limnetic aggregates formed during
48 hours. Results from 0X treatments in which nitrate levels were 0.8 mg
l-1 indicated that ambient lake conditions were sufficient for
the formation of aggregates in February 1997. The addition of only 10 mg
l-1 of total nitrogen, (the upper limit of EPA established safe
water standards), significantly increased the volume but not the numbers
of individual aggregates. Both the individual volume and numbers of aggregates
increased significantly in treatments where the total nitrogen levels exceeded
50 mg l-1.
This increase in aggregate numbers and size may be attributable directly
to physiochemical changes or indirectly to biological (microbial) changes
that resulted from the increase in nutrients. Underlying limestone makes
J. Percy Priest a hard-water lake. Colloid stability and aggregate formation
is dependent on the chemistry of freshwater systems and aggregation typically
is increased in hard water lacustrine systems (Weilenmann et al. 1989).
The addition of nutrients in our treatments may have altered the chemistry
of the water directly and facilitated formation of increased numbers of
larger aggregates. Aggregate formation is also dependent on the release
of exoploymers by bacteria and microalgae (Decho 1990, Kiørboe
& Hansen 1993, Passow et al. 1994). Increased
nutrient concentrations may have affected aggregate formation indirectly
by causing an increase in microalgal growth resulting in the increased release
of exopolymers (e.g.; Kaltenbock & Herndl 1992). Limnetic aggregates
are sites of increased microalgal and bacterial densities (Grossart &
Simon 1993).
The influence of nutrient additions on aggregate size has implications
for particle settlement and the vertical flux of material within lakes.
Large, rapidly sinking aggregates likely are the major source of organic
matter deposition in marine systems (Fowler & Knauer 1986, Alldredge
& Silver 1988). Aggregate physical characteristics will affect the rate
at which particles sink through the water column, with smaller aggregates
typically slowing particle settlement (Alldredge et al. 1990). The increased
size of aggregates in nutrient addition treatments should increase the rate
of water column particle flux and possibly increase the
removal of excess nutrients from the water column.
The increased aggregate density and size with increased nutrient concentrations
demonstrated in our laboratory experiment may have beneficial effects in
natural lake systems. Accelerated surface water removal of excess nutrients
along with pollutants and trace metals that bind to aggregates could improve
overall water quality. Removal of algal blooms through aggregation and flocculation
also might prevent or at least reduce the severity of fish kills often associated
with eutrophic conditions.
Acknowledgments: We thank Dennis Bekaert, Laura Crafton and the other members of the Spring 1997 Marine Biology class for their collaborative efforts in conducting the snow experiment and Jack Ross in Photographic Services for the loan of equipment and photo processing. This research was supported in part by the Department of Biology of Middle Tennessee State University and a grant from NSF Biological Oceanography (OCE 9017807) to A. L. Shanks and K. Walters.
Literature Cited
Alldredge, A.L. and M.W. Silver. 1988. Characteristics, dynamics and significance of marine snow. Progress Oceanography. 20:41-82.
Alldredge, A. L. and M. J. Youngbluth. 1985. The significance of macroscopic aggregates (marine snow) as sites for heterotrophic bacterial production in the mesopelagic zone of the subtropical Atlantic. Deep-Sea Research. 32: 1445-1456.
Alldredge, A. L., T. C. Granata, C. C. Gotschalk, and T. D. Dickey. 1990. The physical strength of marine snow and its implications for particle disaggregation in the ocean. Limnol. Oceanogr. 35: 1415-1428.
Day, R.W. and G.P. Quinn. 1989. Comparisons of treatments after analysis of variance in ecology. Ecol. Monogr. 59: 433-463.
Decho, A. W.. 1990. Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanogr. Mar. Biol. Ann. Rev. 28: 73-153.
Fowler, S. W. and G. A. Knauer. 1986. Role of large particles in the transport of elements and organic compounds through the oceanic water column. Prog. Oceanogr. 16: 147-194.
Grossart, H.P. and M. Simon. 1993. Limnetic macroscopic organic aggregates (lake snow): occurrence, characteristics, and microbial dynamics in Lake Constance. Limnology & Oceanography. 38: 532-546.
Jackson, G.A. 1994. Particle trajectories in a rotating cylinder: implications for aggregation incubations. Deep-Sea Research. 41: 429-437.
Joyner, S.P. 1985. SAS/STAT guide for personal computers. SAS Institute Inc., Carey, NC. ver 6, 378 pp.
Kaltenbock, E. and G.J. Herndl. 1992. Ecology of amorphous aggregations (marine snow) in the Northern Adriatic Sea. IV. Dissolved nutrients and the autotrophic community associated with marine snow. Mar. Ecol. Prog. Ser. 87: 147-159.
Kiørboe, T. and J. L. S. Hansen. 1993. Phytoplankton aggregate formation: observations of patterns and mechanisms of cell sticking and the significance of exopolymeric material. J. Plank. Res. 15: 993-1018.
Lochte, K. and C. M. Turley. 1988. Bacterial and cyanobacteria associated with phytodetritus in the deep sea. Nature. 333: 67-69.
Norusis, M.J. 1993. SPSS for Windows. SPSS Inc., Chicago, IL. ver 6, 828 pp.
Passow, U., A. L. Alldredge, and B. E. Logan. 1994. The role of particulate carbon exudates in the floculation of diatom blooms. Deep-Sea Res. 41: 335-357.
Shanks, A.L. and E.W. Edmondson. 1989. Laboratory-made artificial marine snow: a biological model of the real thing. Marine Biology. 101: 463-470.
Shanks, A.L. and E.W. Edmondson. 1990. The vertical flux of metazoans (holoplankton, meiofauna, and larval invertebrates) due to their association with marine snow. Limnology Oceanography. 35: 455-463.
Simon, M., A. L. Alldredge and F. Azam. 1990. Bacterial carbon dynamics on marine snow. Marine Ecology Progress Series. 65: 205-211.
Weilenmann, U., C.R. O'Melia, and W. Stumm. 1989. Particle transport in lakes: models and measurements. Limnology & Oceanography. 34: 1-18.
Figure Legends
Figure 1: The mean (+ standard error) number of
aggregates per liter (n = 4) formed in roller tanks in which nutrient levels
were manipulated (0X = 0.8 mg l-1, 1X = 10.8 mg l-1,
5X = 50.8 mg l-1).
Figure 2: The mean (+ standard error) volume of individual aggregates (n = 4) formed in roller tanks maintained under varying nutrient treatment levels (see Fig. 1 for levels).
