Studies of nutrient cycling and enrichment in aquatic ecosystems are commonly conductedin enclosed experimental ecosystems. Although there is considerable information about howthe dimensions of natural aquatic ecosystems influence nutrient cycling processes, little is known onhow nutrient cycling studies might be affected by the physical scales of experimental enclosures. Inthe present study, replicate (n = 3) cylindrical containers of 5 dimensions with 3 volumes (0.1, 1.0,10 m3), 3 depths (0.46, 1.0, 2.15 m), and 5 diameters (0.35, 0.52, 1.13, 2.44, 3.57 m) were establishedand subjected to pulsed additions of dissolved inorganic nutrients (DIN, Pod3-, Si) in summer andautumn experiments. Consistent with common experimental protocols, walls of these containerswere not cleaned of periphytic growth during the 8 wk studies. Nutrient concentrations in experimentalecosystems were low prior to nutrient-pulse additions and exhibited exponential depletionfollowing treatments. Overall, larger containers had lower net uptake rates and higher nutrient concentrationsthan did smaller tanks. Relative contributions of planktonic, benthic and wall periphyticcommunities to total nutrient uptake varied in relation to dimensions of experimental systems. Ingeneral, net uptake rates by planktonic communities were inversely related to water depth, withhigher rates associated with increased mean Light-energy in shallower systems. Indirect estimates ofbenthic uptake rates, which were relatively low in all but the shallowest systems, tended also to beinversely related to depth and directly proportional to light levels at the sediment surface. In contrast,nutrient uptake by wall communities (per water volume) was inversely related to the radius of experimentalcontainers. Differences in the 2 container dimensions, depth and radius, accounted for morethan 90 % of the variance in both net nutrient uptake by the whole ecosystem and the molar ratio ofDIN/POd3- concentrations in the water column. Similarly, differences in net nutrient uptake ratesamong experimental ecosystems of different dimensions could be explained by the relative partition-,ing of rates among planktonic, periphytic, and benthic habitats. These results demonstrated that thephysical dimensions of experimental ecosystems can have profound effects on measured nutrientdynamics. We also suggest that many of these experimental observations may be relevant also tomore genera1 scaling relations for nutrient cycling in natural aquatic ecosystems.
