The Influence of Lake and Stream Biogeochemistry on Ecosystem Properties

The Influence of Lake and Stream Biogeochemistry on Ecosystem Properties


Nutrient biogeochemistry affects a myriad of ecosystem processes from primary producer biomass and community composition, to habitat quality, to upper trophic level production. Excessive nutrient loading can cause nuisance algal blooms and hypoxia, and recent studies conclude allochthonous carbon inputs may be important food web subsidies.


Figure 1. The relationships between predicted lake TP concentration and phosphorus retention according to a classic mass balance analysis. This shows our ability to predict lake phosphorus retention is only moderate at best (Brett and Benjamin 2008).

Phosphorus – We used a classic mass balance approach to assess the influence of phosphorus input concentrations and lake hydraulic characteristics on lake TP concentrations. This reassessment of the original Vollenweider type case studies showed the input TP concentration to lakes was by far the best predictor of lake phosphorus concentrations (r2 = 0.71), and that the predict lake TP concentration could further improved by also considering the lake’s hydraulic retention time (Brett and Benjamin 2008). However, despite an apparently good overall fit between predicted and observed TP concentrations (r2 = 0.82), the actually prediction error for individual lakes was ± a factor 4 in log-space. Further, this mass balance approach was only able to predict 35% of the variation in lake phosphorus retention. These results indicate our current understanding of lake phosphorus dynamics is rudimentary at best. The form in which phosphorus is supplied to lakes (i.e. inorganic or organic, dissolved or particulate) probably has a particularly important impact on phosphorus retention (Brett and Benjamin 2008). Phosphorus supplied predominantly as a constituent of animal or human wastes may be readily incorporated into the biotic cycle of lakes. Conversely, phosphorus supplied predominantly in particulate form as P sorbed onto or incorporated into inorganic particles is much less bioavailable to phytoplankton. Particulate, mineral-bound phosphorus can also settle out of the water column rapidly and be lost to the sediments.

Figure 2. The average monthly time series for the total dissolved N (TDN) concentrations and stream flow (upper panels) and TDN export (lower panels) for the mainstem Hood Canal (left panels) and Lynch Cove region (right panels) of this estuary. The error bars presented for the TDN concentrations are ± 1 SD. The various curves presented in the loading panels represent partial TDN loadings attributable to alders, humans and natural background conditions (Steinberg et al. 2011).

Nitrogen – A field study of a watershed discharging to an estuary with recurrent hypoxic and fish-kill events showed both vegetation type and population density were clearly associated with stream nitrogen concentrations. Streams draining catchments with a high prevalence of N-fixing red alders (Alnus rubra) and/or high human population density (and associated onsite septic system discharges) had about 10 times higher nitrogen concentrations than did systems draining the most pristine regions of the watershed (i.e., the Olympic mountains) (Steinberg et al. 2011). Further, there was a very strong seasonality to watershed nitrogen export, with over 60% of the N export from this watershed occurring during the 3 month period of November–January and only 15% occurring during the 5 month period from June to October. Catchments with a high prevalence of alders and high population density had their export dominated by nitrate, whereas catchments with mainly conifer vegetation and low population density exported a high proportion dissolved organic N. However, entrainment of marine water into the surface layer of Hood Canal accounted for ≈ 98% of N loading to the euphotic zone of this estuary. Mass balance calculations indicate red alders and domestic wastewater discharges were a very important N source for many streams, but a minor nutrient source for the estuary as a whole (Steinberg et al. 2011).Carbon – Several studies have concluded that terrestrial carbon inputs support ≈ 50% of zooplankton and fish production in many small-forested lakes, even in cases where these inputs are relatively small compared to phytoplankton production. We conducted a series of experiments to test the food quality of terrestrial resources for zooplankton compared common lake phytoplankton. These experiments showed Daphnia fed diets comprised entirely of terrestrially derived matter had lower growth efficiencies (5 vs. 41%), reproduced later (19.4 vs. 13.5 d), had fewer neonates (3.1 vs. 69.5) and were smaller (0.22 vs. 1.06 mg dry wt. ind.-1) than did Daphnia that consumed phytoplankton (Brett et al. 2009).

Figure 3. Daphnia fed phytoplankton (left) or terrestrial resources (left). These individuals originated from the same female and brood, and were incubated under exactly the same conditions excepting their diets (Brett et al. 2009).

Analyses of Daphnia lipid composition showed they preferentially accumulated algal fatty acids by a factor 10.

Figure 4. The mass influx of dissolved and particulate carbon from terrestrial sources and the in-lake production of bacteria and benthic/pelagic algae based on individual lake observations. Only data from lakes with total phosphorus < 20 µg L-1 were used (Brett et al. 2012).

Very recent studies on this topic have acknowledged the much lower food quality of terrestrial resources, but these studies claimed that the flux of resources from terrestrial sources was so much larger than phytoplankton production that the higher quantity of terrestrially derived resources more than compensated for their lower quality. We compiled all of the data from studies that have directly determined terrestrial inputs and algal production in oligo/mesotrophic lakes. Despite the widespread assumption that terrestrial inputs dominate the carbon flux of many lakes, our analysis indicated algal production is a factor 4-7 greater than the available flux of allochthonous basal resources in low productivity lakes (Brett et al. 2012). Lakes with high t-DOC loading (i.e., 1000 mg m-2 d-1), a median of 98% of the t-DOC flux is advected and therefore is not available to support zooplankton production. When considering both the low quantity and quality of allochthonous resources, our calculations indicate aquatic herbivores are likely to obtain only 2-4% of their resources from terrestrial inputs (Brett et al. 2012). These results indicate terrestrial inputs are likely to make very small direct contributions to animal production in most oligo/mesotrophic lakes.


Brett, M.T. and Benjamin, M.M. 2008. A reassessment of lake phosphorus retention and the nutrient loading concept in limnology. Freshwater Biology 53: 194-211.

Brett, M.T., Arhonditsis, G.B., Chandra, S., Kainz, M.J. 2012 Mass Flux Calculations Show Strong Allochthonous Support of Freshwater Zooplankton Production Is Unlikely. PLoS ONE 7: e39508. doi:10.1371/journal.pone.0039508

Brett, M.T., Kainz, M.J., Taipale, S.J., and Seshan, H. 2009.  Phytoplankton, not allochthonous carbon, sustains herbivorous zooplankton production. Proc. Nat. Acad. Sci. USA 106: 21197-21201.

Steinberg, P.D., Brett, M.T., Bechtold, J.S., Richey, J.E., Porensky, L.E., and Osborne, S.N. 2010.  The influence of watershed characteristics on nitrogen export to and marine fate in Hood Canal, Washington, USA. Biogeochemistry DOI: 10.1007/s10533-010-9521-7.


Michael Brett
(Department of Civil and Environmental Engineering, University of Washington)