FIGURE 2 Seasonal migration timing of Sakhalin taimen during four migration stages (MS) in 2018 and 2019. Circles indicate the observed migration timings of females (red) and males (blue) with lines connecting the same individuals across the years. Spearman’s rank correlation and its significance (*P < 0.05, **P< 0.01, ***P < 0.001) are shown for females (ρF) and males (ρM).
Between-year correlations of the seasonal migration timings were significantly high for female Sakhalin taimen at all MSs (P< 0.05) but were significant only at MS4 (P < 0.05) for males. For the females, within-season correlation between MS1 and MS4 timings was also significant (ρ = 0.628, P < 0.001), but it was extremely low and not significant for the males (ρ = 0.002, P > 0.05). The reason that there are fewer lines for MS4 in Fig. 2 is because we terminated PIT detection before all post-spawners had migrated downstream past Site L in 2018. Only in the females during MS3 was there a significant between-year correlation on the diel-scale as well (i.e. female individuals departed from tributaries at similar times of day in both years) (ρ = 0.497, P= 0.001).
Neither ascent (MS1) nor descent (MS4) by Sakhalin taimen through Site L was uniformly distributed throughout the day (Fig. 3, Rayleigh test,P < 0.001 for females, and P < 0.01 for males). Both female and male spawners ascended the river mostly during daytime, peaking before sunset. Descent by post-spawners peaked early in the afternoon in males, but it peaked during night-time in females. Consequently, the diel migration timing of female Sakhalin taimen differed significantly between ascent and descent (P = 0.000, Watson–Wheeler test), and the timing of descent differed between the sexes (P = 0.003).