Erickson and Nadon: Stepwise stochastic simulation for distributions of missing life history parameter values 
Table 6 
Comparison of the accuracy and precision of estimates of life history parameters 
between use of the stepwise stochastic simulation approach and use of the FishLife 
approach. The reported values are the average results for all 4 of the selected test 
species. The FishLife model generated probability distributions at the genus and 
family levels. The parameters include the asymptotic length (L..), growth coefficient 
(K), natural mortality (4), length at which 50% of individuals are mature (Ly,at), 
and the ratios of L,,,, to L,, and of M to K. The accuracy metric is the standardized 
distance from the median values generated with the meta-analytical approaches 
to values from life history studies (in percentage units). Precision is reported as 
the standard deviation of each parameter (in millimeters for total lengths and on 
an annual basis for M and K). An asterisk (*) indicates that results are the most 
accurate or precise for that life history parameter among the approaches tested. 
Modeling approach L.. K M Lpa tol, Mtok 
Accuracy 
Stepwise 8* 45 11* 
FishLife (genus) 58 30* 99 
FishLife (family) 17 45 86 
Precision 
Stepwise 111* 0.17 0.08* 
FishLife (genus) 449 0.12* 0.14 
FishLife (family) 611 0.21 0.22 
Figure 5 
The 95% confidence intervals for probability distributions of natural mortality (M) and the growth 
coefficient (K) from use of the stepwise stochastic simulation approach (solid circular lines) and 
from use of the FishLife approach at both the genus (dotted circular lines) and family (dashed cir- 
cular lines) levels for the 4 selected test species: (A) javelin grunter (Pomadasys kaakan), (B) red- 
breasted wrasse (Cheilinus fasciatus), (C) camouflage grouper (Epinephelus polyphekadion), and 
(D) blacktip reef shark (Carcharhinidae melanopterus). Published values from life history studies 
(black circles) are also presented for each species, with line segments connecting the median of the 
distributions to the study value. 
87 
