Revisiting Odum (1956): A synthesis of aquatic ecosystem metabolism

H. T. Odum’s influential Limnology and Oceanography 1956 publication compared gross primary production (GPP) and ecosystem respiration (ER) among aquatic ecosystems. Few syntheses of aquatic ecosystem metabolism have been completed since. We used Odum’s conceptual framework to compare GPP and ER from open-water diel oxygen curves in lakes, wetlands, estuaries, and streams (n 5 350). We also documented environmental drivers of metabolism among ecosystems. GPP and ER were strongly related in lakes and estuaries, but weakly related in streams and wetlands. GPP and ER were highest in estuaries, and GPP : ER was lowest in streams. Differences in the magnitude and variability of metabolism among ecosystems were attributable to landscape and water-column factors. Watershed size and phosphorus (P) concentrations were positively related to GPP and ER across all ecosystems. Considered independently, lake and estuary GPP were driven by P concentrations. In contrast, land-use and canopy cover drove stream metabolism, not nutrient concentrations. Results confirmed the classic paradigm that estuaries are the most productive aquatic ecosystem; however, our synthesis showed that relative to streams and estuaries, there was higher variation in lake GPP and ER than previously documented. Results will be valuable for management, restoration, and carbon budgets, which incorporate metabolism measurements at both the catchment and landscape scales. As metabolism datasets grow, future syntheses will address challenges including seasonality, sensor deployment time and location, hydrology, and variation in analytical conventions by discipline. Ongoing technological and computational advancements, combined with increased communication among subdisciplines, should also expand insights generated by subsequent metabolism syntheses. Gross primary production (GPP), ecosystem respiration (ER), and net ecosystem production (NEP) are fundamental metrics of an ecosystem (Odum 1956; Marcarelli et al. 2011; Staehr et al. 2012b). GPP is the rate of organic matter production within an ecosystem by photosynthesis, whereas ER (also referred to as community respiration) represents the total consumption of organic matter in an ecosystem via aerobic respiration. NEP, also referred to as net ecosystem metabolism, is the balance between GPP and ER. Because metabolism metrics integrate the activity of all organisms carrying out photosynthesis and aerobic respiration, they can be compared across locations and through time to infer how entire ecosystems respond to environmental change (Odum 1956; Mulholland et al. 2001; Caffrey 2004). Models of global carbon (C) sinks and sources are based upon syntheses of empirical studies that measure metabolism across a diversity of ecosystem types and conditions including aquatic (Aufdenkampe et al. 2011) and terrestrial environments (Yvon-Durocher et al. 2012). Therefore, it is critical to document the balance between autotrophy and heterotrophy, as well as compare controls of GPP and ER across ecosystems. Ecosystem metabolism is commonly measured in all aquatic ecosystems, including streams (Young et al. 2008; Bernot et al. 2010), lakes (Cole et al. 2000; Staehr and Sand-Jensen 2007), wetlands (Hagerthey et al. 2010), and estuaries (Howarth et al. 1991; Caffrey 2004; Russell et al. 2006). However, in half a century of research on aquatic ecosystem metabolism, empirical analyses among disparate ecosystem types have rarely extended beyond comparing within each ecosystem type across a geographic space (Caffrey 2004; Bernot et al. 2010), through time (Roberts et al. 2007), or among varied trophic states (Staehr et al. 2010). A recent meta-analysis of research foci for publications that included aquatic ecosystem metabolism showed that although an equal proportion of studies were conducted in rivers, lakes, estuaries, or the coastal ocean between 2000 and 2010, only , 3% of studies considered more than one type of ecosystem (Staehr et al. 2012b). Past data syntheses have assessed a combination of terrestrial and aquatic environments (Whittaker and Likens 1973; Begon et al. 1986; Yvon-Durocher et al. 2012), but none have focused specifically on aquatic ecosystems. Expanding metabolism analyses to include multiple ecosystem types could illustrate patterns in the relative ranges of GPP and ER among ecosystems and reveal environmental drivers of metabolism that may be obscured when ecosystem types are considered independently. Partitioning aquatic metabolism research by ecosystem is contrary to a foundational paper on the subject. Odum (1956) used diel oxygen (O2) curves to compare rates and environmental drivers of GPP and ER in flowing waters relative to other aquatic ecosystems. This paper provided an analysis that summarized both the absolute and relative values for GPP and ER across aquatic ecosystems in a single figure (Fig. 1A). Odum (1956) showed that the magnitude and the relative values of GPP and ER varied between standing and flowing waters, in rivers receiving terrestrial or anthropogenic organic C inputs, and in * Corresponding author: [email protected] Limnol. Oceanogr., 58(6), 2013, 2089–2100 E 2013, by the Association for the Sciences of Limnology and Oceanography, Inc. doi:10.4319/lo.2013.58.6.2089