Temporally-variable productivity quotients on a coral atoll: Implications for estimates of reef metabolism

Measurements of net ecosystem productivity (NEP) and calcification (NEC) from contemporary coral reefs provide a baseline for monitoring the impacts of future stressors like ocean acidification and sea-surface warming. However, separating secular trends from natural variability requires NEP and NEC records across a wide range of spatial and temporal scales. One promising way to make these measurements is with autonomous pH and O-2 sensors. Crucially, the accuracy of this approach relies on knowledge of the in situ ecosystem productivity quotient (Q), which indicates the moles of O-2 consumed per mole CO2 produced. Using co-located measurements of total dissolved inorganic carbon (DIC), total alkalinity, and dissolved O-2, we empirically determined Q during a three-year field campaign on Tetiaroa Atoll, French Polynesia. Empirical values of daily net Q (-1.02 <= Q(net) <= -0.47) frequently differed from both the canonical value of -1.45 for the ocean and the value of -1 often assumed for tropical reef ecosystems. Furthermore, Q changed on hourly timescales, and integrated daily values differed between days. We hypothesize that captive bubbles on the surfaces of coral, macroalgae and other substrates can explain these variations in Q, with other influential mechanisms being mixing between parcels of water, sedimentary denitrification, and ammonium-fueled primary productivity. Our findings, which are robust to changes in the model that is used to correct for advection and gas exchange, as well as changes in model parameters, suggest that future investigations of metabolism on reefs should be based on measurements of carbonate chemistry variability, rather than O-2 evolution. Otherwise, large biases in NEP and NEC could result (> 40%).