The exponent values of the length-mass relationships of both C. heterodon (3.17) and C. guichenoti (2.93) estimated in this study (Figure 1) were within the general range found for fishes (2.5 to 3.5) in previous studies (Carlander 1969). The exponent value of C. heterodon indicates a positively allometric growth, while the exponent value of C. guichenoti (2.93) indicates an approximately isometric growth. The faster increase of weight than length of C. heterodon in this study indicates that bone mass tends to increase at a slower rate than does muscle mass, as stated by Shearer (1994). Correspondingly, in this study ASH decreased with M of C. heterodon (Figure 4), which could reflect a slower increase in bone mass. The isometric growth of C. guichenoti is also consistent with its unchanged ASH with M.
FAT of C. guichenoti was relatively higher than that of C. heterodon (Figure 2). Childs and King (1993) classified fishes into low fat fish (FAT within 0.6-3.0%), intermediate fat fish (FAT within 3.5% to 7.0%) and high fat fish (FAT within 8.5% to 15.3%). C. guichenoti (FAT ranges 5.66% to 20.4%) could be incorporated into the intermediate or high fat categories and some individuals of larger size even having FAT above the range of high fat categories. C. heterodon (FAT ranges 2.02% to 14.6%) could be incorporated into intermediate or low fat categories. It has been found that the long distance migratory species deposit larger amounts of body lipid (Jonsson and Jonsson 2005), which could be the reason for the higher FAT of the migratory C. guichenoti than that of the residential C. heterodon.
FAT, PRO, and E of many fish species increase as the body size increases (Berg and Bremset 1998; Jonsson and Jonsson 1998, 2003; Sogard and Spencer 2004). Similar results were also found in the both species in this study (Figures 2, 3 and 5). However, FAT of C. guichenoti showed an interesting pattern of increase, where FAT increased rapidly and then leveled off as M increased above 6.5 g. E of C. guichenoti also increased in a similar two-line pattern but with a relative larger transition body mass of 19.1 g. This was partly contributed to by the persistent increase in PRO. This suggests that the lipid concentration of C. guichenoti reaches an upper limit. Thus, the energy storage of larger fish needed for migration may depend mainly on increase of body size rather than body energy density.
PRO of the smaller individuals of both species was higher than FAT, indicating that more intake energy was allocated to synthesis of protein than lipid (Figures 2 and 3). Similar results were also reported in Atlantic salmon (Salmo salar), brown trout (S. trutta), and sablefish () (Berg and BremsetAnoplopoma fimbria1998; Jonsson and Jonsson 1998; Sogard and Spencer 2004). The strategy of protein build-up of fish fry might enhance their competition capacity and reduce predation risk (Calow 1985). Furthermore, FAT of both species increases faster than PRO as the fishes grow. This suggests that the importance of lipid increases and there is a shift in energy allocation from protein synthesis to lipid storage. Similar results were also found in gulf menhaden () (DeeganBrevoortia patronus1986) and many other species (Shearer et al. 1994; Anthony et al. 2000).
Previous studies (Hartman and Brandt, 1995; Jonsson and Jonsson 1998; Pangle and Sutton 2005; Hartman and Margraf 2008) have suggested using dry mass content or water content to predict the proximate composition of fishes. The present study also found significant relationships between WAT and FAT, PRO, ASH, or E in both fish species (Figure 7 a, b, c, d). The models for FAT or E of both species yielded high r2 values (range: 0.731 to 0.964) suggesting strong predictive power for future application. However, variation of PRO and ASH could be explained less well by WAT (r2 values range: 0.054 to 0.237). The lower predictive power of models for PRO and ASH may be related to narrow ranges of protein and ash contents in both C. heterodon and C. guichenoti. Fishes may exchange body water and fat when energy budgets change (Hartman and Margraf 2008). Previous studies suggest that the equal amounts of decrease in body water are associated with the accumulation of around three times as much lipid as protein (Schmidt-Nielsen 1975; Jobling 1994). The present results showed that the slopes of WAT-FAT model were 4.1 times higher in C. heterodon and 2.3 times higher in C. guichenoti, compared with the slopes of WAT-PRO model, suggesting a rapider accumulation of lipid. Similar rapider accumulation of lipid was also found in brown trout (Jonsson and Jonsson 1998) and Atlantic salmon (Jonsson and Jonsson 2003).
Our results also showed significant differences between the two species of the slopes in both WAT-FAT and WAT-E models (Figure 7 a, b). Equal changes in body water would induce 1.4 times the change in FAT and 1.3 times the change in E for C. guichenoti compared with C. heterodon. This suggests that, even between closely related species, FAT and E cannot be predicted by WAT using general models. Further work is needed to determine whether the stronger replacement between water and lipid in C. guichenoti is related to its migratory characteristics.
For future work, different models for the two species should be used to predict FAT or E by WAT. The migratory C. guichenoti has a higher FAT than that of the residential C. heterodon. With dam constructions in the upstream region of the Yangtze River, C. guichenoti is undergoing loss of its migratory pathway and even its migratory behavior. Its energetic response to the intense changes of habitat remains unclear and would be an interesting area of future research.