Table 1.
Differences between traditional and modified truncated cones methods.
Fig 1.
Morphometric measurement locations.
Locations of width, height, girth, and curvilinear measurements. Dorsal and lateral ultrasound images were also taken at all locations except ears and ankles. Star indicates location of blubber biopsy. Photo by Patrick Robinson under NMFS Permit 14636.
Fig 2.
Lateral position at the sternum location along the length of the animal (see Fig 1), late-molt ultrasound image of blubber and skin layers. Peaks represent compressions.
Fig 3.
Width vs. height at different locations along an elephant seal’s body (N = 14 animals, each animal independently measured three times). Points along the diagonal red line indicate circularity.
Fig 4.
Difference in sculp ultrasound thickness comparing ultrasounds with and without an image for dorsal and lateral surfaces of the seals from pelvis to neck.
Points are means, and bars represent ± 1 standard deviation combining distributions of triplicate measures for dorsal surfaces and six measures for lateral surfaces from 13 seals.
Table 2.
Posterior marginal distributions of parameters relating sculp depth ultrasound measurements from images (YU,i) to ultrasound measurements without images (XU,i) at different locations (i): pelvis, umbilicus, mid, sternum, axilla, and neck, on the dorsal and lateral surfaces.
Fig 5.
Comparison of truncated cones techniques to estimate blubber volume.
Traditional truncated cones technique (circular cones without ultrasound images, not accounting for skin, blubber from neck to pelvis) vs. a new method (elliptical cones with ultrasound images, accounting for skin, blubber from ears to ankles). Points are means and circles are 95% posterior intervals. Dashed black line is parity. Orange lines are mean (solid) and 95% posterior interval (dashed) values from a hierarchical Bayesian analysis accounting for uncertainty in measurements.
Fig 6.
Proportion fat vs. proportion total body water.
Relationship between proportion total body water as measured from tritiated water dilution and proportion fat from dissection. Points are data for three pinniped species. Solid black line is mean and dashed lines are 95% posterior intervals from results of Eq 6. The orange line represents the estimated linear relationship without uncertainty by Iverson et al. [26]: MassFat = MassTotal− 1.37MassWater. The solid blue line is the estimated mean value of proportion fat given proportion water from guinea pig carcasses [19] with 95% posterior intervals (dashed blue lines). For posterior distributions of Eq 6 parameters, see Results. For the guinea pig equation derivation and R code, see S1 File and S3 Code.
Fig 7.
Comparison of techniques to measure proportion adipose tissue and proportion fat.
Proportion adipose tissue (top) and proportion fat (bottom), showing results from the traditional truncated cones method (left) and the modified cones method (right) compared to labeled water estimates. Points are means, and circles represent 95% posterior range. Dashed diagonal line indicates a 1:1 relationship. Red shaded area with red line indicate mean and 95% confidence interval for juveniles reported by Webb et al. [21] using the tritiated water method. Gray shaded area with red line indicate mean and 95% posterior interval for molting adult females reported by Worthy et al. [24] using the tritiated water method.
Fig 8.
Standard deviation of total proportion fat when measurements are independently varied using the modified truncated cones method. Height (H), width (W), dorsal ultrasound measurement (USD), and lateral ultrasound measurement (USL) refer to elliptical truncated cone measurements. CL = curvilinear length segment, SDepth = skin depth, DensB = density of blubber, FLB = proportion fat in blubber, NCL = curvilinear length from ears to tip of nose. Points are means and bars are standard deviations for all measured animals (N = 14).
Fig 9.
Standard deviation of total proportion fat from tritiated water method when measurements are independently varied. Water-to-fat is the uncertainty associated with the relationship between proportion water and proportion fat, as measured in Eq 6 with results in Fig 6. CPMstock is derived from standard solutions. CPM conc bg is the background CPM·mL−1, and CPM conc equil is the CPM·mL−1 after the labeled water had equilibrated in the animal. Points are means and bars are standard deviations for all measured animals (N = 10).