Pella and Masuda: Bayesian methods for analysis of stock mixtures from genetic markers 
163 
Iteration 
Figure 3 
The three chains of posterior samples for the unknown proportion from the upper Sashin 
Creek steelhead population present in the lower Sashin Creek stock mixture, based on 
microsatellite data. Each chain was initiated with a different value for the unknown upper 
proportion, 0.95, 0.50, and 0.05. All chains tracked to high proportions during the early 
phase and later stabilized at the equilibrium posterior distribution. 
Table 4 
Parameters of the marginal posterior density for the upriver steelhead proportion in the lower-river stock mixture. 
Posterior quantiles 
Data 
Mean 
Mode 
SD 
2.5% 
Median 
97.5% 
Allozymes 
0.25 
0.22 
0.07 
0.12 
0.24 
0.40 
Microsatellites 
0.24 
0.22 
0.06 
0.13 
0.24 
0.37 
MtDNA 
0.41 
0.00 
0.19 
0.01 
0.46 
0.70 
The posterior marginal distributions of the upriver propor- 
tion from allozymes and microsatellites were nearly identical 
and their modes (0.22) or means (0.24, 0.25) were support- 
ive of the earlier approximate assessment of 0.25 (Table 4). 
Either data set points with high probability (0.95) to an up- 
river population presence between 0.1 and 0.4 of the lower- 
river stock mixture. The mtDNA was not nearly as informa- 
tive in regard to population composition, judging from the 
