Fig. 4. Effect of hypoxic gas injection on egg-chamber P_O2 (A,C) and estimated air-adding frequency (B,D) of two male Periophthalmus modestus. (A,B) Burrow instrumented on June 30, 2004. (A) Real-time records of the low-tide increase in egg-chamber P_O2 contrasted with the significantly higher (3.2x, P<0.0001, ANCOVA) P_O2 increase rate observed over 60 min following hypoxia injection (arrowhead) than at the beginning of the low-tide period with nearly identical initial P_O2. (B) An inverse relationship between egg-chamber P_O2 and air-adding frequency (f_a) as determined for five complete low-tide periods preceding hypoxic-gas injection (solid symbols) and f_a after injection (open symbol). Egg-chamber P_O2 at the beginning of each 60 min segment is plotted against f_a (see Appendix). Regression line equation; f_a=1.39P_O2^–0.872 (P=0.006, r=0.449, the post-hypoxia value is not included in the regression). (C,D) Burrow instrumented on July 29, 2004. (C) Real-time records of egg-chamber P_O2. Hypoxia injection (arrowhead) caused 3.2-fold higher P_O2 increase than at the beginning of the low-tide period (P<0.0001, ANCOVA). (D) An inverse relationship determined for four low-tide periods and higher f_a immediately following hypoxia injection, as in A. Regression line equation; f_a=19.68P_O2^–1.729 (P<0.0001, r=0.881). Solid and open symbols represent data points obtained before and after hypoxic-gas injection, respectively, as in B.