Original paper

Alternative approaches for estimating components of lake metabolism using the free-water dissolved-oxygen (FWDO) method

McNair, James N.; Sesselmann, Meagan R.; Kendall, Scott T.; Gereaux, Leon C.; Weinke, Anthony D.; Biddanda, Bopaiah A.

Fundamental and Applied Limnology Volume 186 Nr. 1-2 (2015), p. 21 - 44

published: Feb 1, 2015

DOI: 10.1127/fal/2015/0626

BibTeX file

ArtNo. ESP141018601002, Price: 29.00 €

Download preview PDF Buy as PDF


The free-water dissolved-oxygen (FWDO) method employs high-frequency time series of dissolved oxygen (DO) and several other variables to estimate gross primary production (GPP), total respiration (R), and net production (NP) of a lake. It requires a process-based model of DO dynamics to represent the roles of oxygenic photosynthesis, aerobic respiration, and atmospheric exchange in determining observed DO concentrations, as well as a procedure for estimating parameter values. We review approaches that have been used in conjunction with the FWDO method for modeling DO dynamics in lakes and estimating model parameters. Most studies have used mixed-layer models of DO dynamics that assume stratified lakes have a well-mixed upper layer (mixed layer) whose DO concentration is not affected by deeper water, which therefore is excluded from the models. We have found that mixed-layer models often cannot adequately account for observed diel dynamics of DO in our study system (Muskegon Lake, USA), indicating that model improvement is necessary. We outline various alternative forms of mixed-layer DO models that are compatible with the FWDO method (including several new ones), review methods for estimating GPP, R, and NP with them, and illustrate their performance in Muskegon Lake. None of the alternative forms resolve the problem of frequent poor model fits. It therefore appears that an important process is missing from mixed-layer DO models. We suggest the main problem is that the mixed-layer assumption is too simple to adequately represent the key physical transport and mixing processes in Muskegon Lake on most days, and we argue that this problem is likely to arise in other lakes, as well. We conclude by discussing several alternatives to mixed-layer models that address the entire water column, are straightforward to implement, and retain compatibility with the FWDO method.


high-frequency time series of dissolved oxygennet productiontotal respirationdiel dynamics of dissolved oxygenmuskegon lakegross primary production