Fig 1.
First phase (top): Four levels of one treatment, feeding rate (12.5, 25, 50, 100%). Second phase (bottom): Four levels of the treatment salinity (0, 8, 16, 24/32 ppt), and three levels of the exposure time to salinity levels (12, 72, 120 h). The experimental design was the same for the two species, except that maximum salinity was different for white (W) and green (G) sturgeon.
Fig 2.
Standardized values (0 to 1) of biological responses to nutritional changes.
Green and white dots denote green and white sturgeon, respectively. For each biological parameter (y-axis), fishes were challenged with four levels of feeding rate (12.5, 25, 50, 100%; right y-axis). Three observations were collected per FR level per species.
Table 1.
Significance levels of main effects and interactions for feeding rate and species on biological responses.
Fig 3.
Generalized additive model fits and 95% confidence intervals from the optimal model for green sturgeon.
Relationships between growth performance of juvenile green sturgeon, measured by condition factor, and each explanatory variable (FR × salinity interaction, time × salinity interaction, pyloric ceca NKA, and muscle moisture). Sturgeons were challenged in a multiple-factor setting, first to levels of feeding rate (12.5, 25, 50, 100%), and then to salinity levels (0, 8, 16, 32 ppt) for a given time (12, 72, 120 h). W = weight; L = length.
Table 2.
Significance of final models’ terms.
Fig 4.
Generalized additive model fits and 95% confidence intervals from the optimal model for white sturgeon.
Relationships between growth performance, measured by condition factor, and each explanatory variable are illustrated. Sturgeons were challenged in a multiple-factor setting, first to levels of feeding rate (12.5, 25, 50, 100%), and then to salinity levels (0, 8, 16, 24 ppt) for a given time (12, 72, 120 h). Only feeding rate and time were kept in the final model. W = weight; L = length.
Fig 5.
Fitted structural equation model for green sturgeon.
Bayesian estimation was used to determine variables affecting condition factor (double-line rectangle). Solid rectangles: treatments; dashed rectangles: osmoregulation parameters. Arrows represent causal pathways between variables, each having a standardized partial regression coefficient (sign indicates whether the relationship is positive or negative for that direct effect).
Fig 6.
Schematic overview of the primary results.
The main relationships among the principal variables identified in this study are summarized as affecting directly or indirectly the growth performance of green and white juvenile sturgeon when challenged with levels of their feeding rate, salinity, and exposure time to salinity. Solid circles: treatments; dashed circles: osmoregulation parameters; double-line circles: growth performance measured by condition factor (CF). The sign associated with each arrow indicates whether the relationship is positive or negative for that effect. Wider arrows refer to the most significant relationships in this study.