Analysis of morphological variation
The shape of the specimens was described using relative warp analysis (Table S2). The first relative warp (RW1) shows the length of the snout and gill cover, and the second relative warp (RW2) explains the angle of the snout (Table 2). The first three relative warps explained 65.91% of the variance in RWs, and the ANOVA results showed that there are significant differences between populations in terms of RW3 (the third relative warp), which represents eye size (Table 2). Specimens plotted on the positive axis of RW3 tend to have larger eye, whereas specimens plotted on the negative axis have smaller eyes (Figure 1b,c).
The morphological difference in different pipefish populations shown by the RW scatter plot reveals that the YS population generally has a higher RW3, which indicates that the fish in this population have larger eyes. In comparison, individuals in the SCS population have lower RW3 values than those in the other two populations, with ECS population fish having intermediate RW3 values (Figure 1d). We found that seven of the total 26 landmarks were related to morphological variation in eye size, which accounts for 51.4% of the variation in information determined by RW analysis (Figure 1e).