Figure 3: Superoxide concentrations across adjoining M. capitata and P. compressa.

(a) Close-up photo of adjoining M. capitata (left) and P. compressa (right). (b) Continuous superoxide measurements were made by moving the sample tubing across the coral surfaces, pausing for 20–60 s at the following positions: along the M. capitata surface at distances of (a) 10 cm, (b) 5 cm, (c) 1 cm and (d) <1 cm (shaded area) from the species interface; and along the P. compressa surface at distances of (e) <1 cm, (f) 1 cm, (g) 5 cm and (h) 10 cm away from the species interface. Chemiluminescence signals were converted to superoxide concentrations by first subtracting out signals of an aged filtered seawater baseline (not shown), and then corrected signals were converted to concentration using the daily calibration curve. The superoxide concentrations shown for the corals include both the coral-derived signal and the seawater (SW) signal (not SW corrected on the trace). Superoxide concentrations for M. capitata (averaged from a to c) and P. compressa (averaged from e to h) were significantly different (P=0.014) based on a two-sample t-test and are indicated below the superoxide trace, along with average superoxide levels in the background SW adjacent to each coral. On the basis of a two-sample t-test, superoxide levels at the M. capitata surface (a–c) were statistically similar (P>0.1) to the levels in the background SW, except right at the edge of the species interface (d, shaded area). Superoxide concentrations are reported as average±s.d. of the temporal signal.