International Journal of Marine Science, 2017, Vol.7, No.5, 37-50
44
different between the Red Sea and Mediterranean Sea fish populations.
2 Results
The correlation between the transformed morphometric distances and total length were found non-significant
(p>0.001), indicating that the effect of size has been successfully removed by performing the allometric
transformation (Equation 1). The PCA on morphometric distances generated the proportion of variation explained
by each principal component (PC). These proportions were assessed using a scree plot to determine the number of
PCs that has to be retained for further evaluation and interpretation (Figure 3). The scree plot of these proportions
indicates that the first three components are large enough to explain a cumulative of 69.22% of the total
morphometric variation. However, only the first two components were considered in this study since the third
component did not show any significant improvement with the separation of population clusters.
Differentiation of populations based on location and sex
The first principal component (PC1) explained 38% of the total morphometric variation (Figure 3) and loaded
heavily on distances representing the tail (CPL and CPD) and eye (ED) of the fish (Table 2). The second principal
component (PC2) explained 19% of the morphometric variation (Figure 3) and loaded heavily on distances
representing the head (HL and HH) and snout of the fish (SnL, Table 2 and Table 3). A bivariate plot of PC1 vs.
PC2 scores of all individuals showed that the samples from Red Sea and Mediterranean Sea are morphologically
distinct giving two separate clusters (Figure 4a). Morphological analysis is also efficient to determine shape
differences associated to sexual dimorphism of the species. However, no such differences (or clusters) were
observed in our analysis when the PC scores of the two sexes were used in the bivariate plot (Figure 4b). Fish
from the Medditerranean stock had smaller heads and snouts in relation to their bodies than those from the Red
Sea.
Figure 3 Scree plot indicating the proportion of variation explained by each principal component in the PCA
Note: The first two components cumulatively explained 57% of the total morphometric variation
The distances related to tail and eye (that loaded heavily on PC1) alone explained most of the shape differences in
the morphology between the two populations. This can be attributed to an adaptation of locomotion and vision of
the fishes, thus they may relate to the swimming or navigational abilities in their respective environments. The
shape differences related to head and snout (that loaded heavily on PC2) are less significant when compared to
those loaded on PC1, but can be attributed to their feeding adaptations in these environments. We discuss these
inferences based on the differences in ecological environment of both habitats along with other evidences on the
phenotypic plasticity of fish species.
