Because of their manageable size, only approximately 20% of find more 1 and 2 yr

old seals were sedated when measured, compared to over 50% of 3 yr olds, and over 90% of seals age 4 yr and older. Dorsal straight length (from tip of nose to tip of tail) and axillary girth (just posterior to the insertion of the foreflippers on inhalation) were measured. We analyzed measurements of live, free-ranging monk seals. Dead seals were excluded as were measurements of seals brought into captivity for rehabilitation or permanent care during the calendar years when held captive. When more than one measurement was available for an individual in a given year, we selected the most complete set of measurements

(i.e., simultaneous length and girth) and, secondarily, the measurements taken closest to 30 June (mid-year). Measurements from different dates in the same year were combined if, for example, length was taken on one date and girth on another. Repeat measurements of the same seal in different years were treated as independent. Integer ages were assigned and incremented on 1 January. We further limited our analysis to seals whose ages were known from marking with flipper tags in the birth year. Measurements from seals aged 1 yr and older were analyzed, thereby excluding immediate postweaning measurements, to best characterize lifetime growth following the period of maternal care. Consistent with McLaren’s (1993) summary of growth in pinnipeds, selleck kinase inhibitor we used the generalized von Bertalanffy function (Schnute 1981) selleck for growth in length: where Lx is length at age x, L∞ is the asymptote, a determines the rate of approach to the asymptote and b determines the curvilinearity of that approach. The parameter x0 is the time before birth when

length equals zero. Rather than attempt to fit x0 we instead assigned a fixed value. Johanos et al. (1994) estimated the mean time from mating until birth in Hawaiian monk seals to be 330 d, or 0.9 yr. However, many pinnipeds have delayed implantation and it is not known whether this occurs in Hawaiian monk seals. Any such delay would reduce x0 below 0.9 yr. Consistent with McLaren (1993) we fixed x0 at 0.73 (approximately 9 mo). This age-offset parameter has little effect on the lifetime growth curve when it is small relative to the life span. The same von Bertalanffy function was used to fit growth in axillary girth. Curves were fitted using the nls (nonlinear least squares) function in R (R Development Core Team 2009). Typically, three parameters (L∞, a and b) were estimated. We evaluated the influence of sex and subpopulation by comparing the small sample Akaike information criterion (AICc) of competing models. Sample sizes were insufficient to evaluate temporal variability in growth; all years were combined in analyses.