Prenatal data impacts common bottlenose dolphin (Tursiops truncatus) growth parameters estimated by length‐at‐age curves

Abstract

Compilation of marine mammal demographic data is central to management efforts. However, marine mammal length-at-age growth curves demonstrate limitations. Physiological growth parameters of terrestrial mammals are typically estimated using curvilinear models fit to size-at-age data along a time series from conception to senescence. The difficulty of collecting and aging prenatal cetaceans is addressed here, and growth parameters of common bottlenose dolphins (Tursiops truncatus) along coastal Texas were estimated using length-at-age information from a broader scope of age classes, including late-term fetuses. A Gompertz growth curve fit to pre-and postnatal data underestimated size parameters, but demonstrated similar growth rate constants (k) to an exclusively postnatal model. However, when growth parameters were broken out, the absolute growth rate (G) and rate of growth 195 196 MARINE MAMMAL SCIENCE, VOL. 27, NO. 1, 2011 decay (g) decreased (0.44 from 0.27 and 0.55 from 0.39, respectively), which underscores the importance of reporting k in its expanded form (G/g). Although the Gompertz fits most age classes well, it cannot explain growth in all age classes. We argue that a novel sigmoidal model would be more useful for inference.

Early physical maturation of female common bottlenose dolphin Tursiops truncatus in the eastern Levantine Basin

“…Size sexual dimorphism of western North Atlantic coastal CBD, with males being slightly larger, was described by almost all investigators employing large enough series (Sergeant 1973;Read et al 1993;Tolley et al 1995;Fernandez & Hohn 1998;Stolen et al 2002;Mattson et al 2006;Turner et al 2006;McFee et al 2010;Neuenhoff et al 2011). Asymptotic lengths of males in the southern US Atlantic coast and Gulf of Mexico range between 10 and 20cm (4-8%) higher than those of females, and this relationship seems to hold for the population hereby reported.…”

“…The first half-year of life of CBD has been suggested to be a separate growth stage with its own parameters (Kasuya et al 1986;McFee et al 2010) or with parameters common to late uterine growth (Neuenhoff et al 2011). This perinatal period is characterized by a high growth rate resulting in a 1.7-1.75-fold increase in body length (as predicted by both Gompertz and von Bertalanffy growth models).…”

“…Goodness of fit of all curves, acquired by least squares iterations, was judged by R 2 values, residual sum of squares and Akaike's Information Criterion (AIC) values. Correlation with regional primary productivity To this end, we could only find four other studies that fitted single Gompertz curves to a point database from dead CBD and reported G and g constants for both genders: three from the coast of Texas (Fernandez & Hohn 1998;Turner et al 2006;Neuenhoff et al 2011) and one from the Indian River Lagoon, Florida (Stolen et al 2002). The values of the first three studies were averaged and the three regional sets of male and female G and g values were log transformed and correlated (Pearson's product correlation coefficient) to corresponding log transformed regional mean annual primary productivity rates obtained from the literature.…”

“…Growth curve fitting Zullinger et al (1984) report that among sigmoidal growth models, the Gompertz model provides good estimates for the largest number of mammalian species, and, indeed, it has been used to describe the growth of many toothed whales (Perrin et al 1976(Perrin et al , 1977Bloch et al 1993;Ferrero & Walker 1995Marsili et al 1997;Siciliano et al 2007) as well as several populations of CBD on the eastern coasts of the United States (Read et al 1993;Fernandez & Hohn 1998;Stolen et al 2002;Neuenhoff et al 2011). In this study, von Bertalanffy as well as Gompertz growth models were fitted to length-atage data.…”

“…Most authors who reported on smaller samples used a single sigmoid curve for the entire sample (i.e. Stolen et al 2002;Turner et al 2006;Neuenhoff et al 2011), some (Read et al 1993;Fernandez & Hohn, 1998) avoiding the stage 1 issue by initiating the curves at the age of one year, and others (Neuenhoff et al 2011) acknowledging the fact that a single-stage model cannot account for the high growth rate observed in young dolphins, particularly when late fetal growth was considered. Mattson et al (2006) applied two-staged sigmoid curves with the intersection points at the age of about eight years for CBD from Mississippi Sound and got satisfactory results; however, they too had to exclude non-neonate animals younger than one year for part of the calculations.…”