sampling error. In order to assess the likelihood the population is non-hybridized we determine the chances of 

 not detecting as little as a one percent genetic contribution of a non-native taxon to a hybrid swarm. This is 

 simply 0.99 "^'"^ where N is the number offish in the sample and X is the number of marker loci where the/? allele 

 is characteristic of the non-native taxon. 



In samples showing evidence of hybridization, that is; fragments characteristic of a non-native taxon were 

 detected at two or more marker loci, we used two approaches to determine if the population appeared to be a 

 hybrid swarm. First, contingency table chi-square analysis was used to test for heterogeneity of allele frequencies 

 among the marker loci. Next, we compared the observed distribution of the number of loci per individual at 

 which non-native fragments were detected to the expected random binomial distribution based on the estimated 

 native and non-native genetic contributions to the population. If both analyses were non-significant we 

 concluded the population came from a hybrid swarm. , , ... 



Heterogeneity of allele frequencies among marker loci can arise in very old hybrid swarms as the frequencies 

 over time diverge from each other due to genetic drift. In this case, however, the non-native fragments will still be 

 randomly distributed among individuals. 



There are two likely reasons why a non-random distribution of non-native fragments may be observed among 

 individuals in a sample. It may contain individuals from genetically divergent populations with different amounts 

 of hybridization or hybridization may have only recently occurred in the population. Based on genetic data alone, 

 these two situations will generally be difficult to distinguish from each other. Regardless of the explanation, 

 when the non-native fragments are not randomly distributed among individuals in a sample estimating a mean 

 level of hybridization has little, if any, biological meaning and, therefore, is often not estimated. 



I' Results and Discussion 



Cabin Creek 2977 



PINE fragments characteristic of westslope cutthroat and rainbow trout were detected at three of the six diagnostic 

 markers between these fishes that were analyzed in the sample. Although the frequency of rainbow trout alleles 

 among the marker loci was statistically homogeneous (P>0.50), the markers characteristic of rainbow trout were not 

 randomly distributed (P<0.001) among the fish in the sample. In contrast, all the rainbow trout markers were detected 

 in only one fish. All the other fish possessed markers characteristic of only westslope cutthroat trout. These 

 results suggest this population is a mixture of hybridized and non-hybridized westslope cutthroat trout. The vast 

 predominance of what appear to be non-hybridized westslope cutthroat trout in the sample and the relatively high 

 frequency of rainbow trout fragments (0.30) in the fish definitely of hybrid origin suggest the latter is a recent migrant 

 into the population. Thus, interbreeding between non-hybridized westslope cutthroat trout and fish of hybrid origin 

 may not have begun at the time of sampling. The presence of fish of hybrid origin in the population, however, 

 seriously ihreaiens the continued genetic integrity of the non-hybridized fish. 



Sincerely, 



Ben Wright 

 Robb Leary 



t.i ■. /I '* 



