YCT Multi-state Assessment February 10, 2003 



where no genetic testing had been done (Table 1 ). Genetic sampling involves many complex 

 issues that can make clear interpretation and reporting of results difficult, especially within 

 standardized databases. We will briefly address a few of these issues here, but suggest reading 

 Appendix D in Shepard et al. 2003 for more detail. 



Table 1 . Genetic classes used for assessing genetic status of Yellowstone cutthroat trout in 2001 . 



Code Description Genetic Report Code 



A Genetically unaltered (<1% introgression) - tested via Tested; Unaltered 



electrophoresis or DNA 



B Introgressed 75% or less and 99% or more - tested via Tested; <=25% to >=1% 



electrophoresis or DNA introgressed 



C Introgressed more than 75% - tested via electrophoresis or Tested; > 25% introgressed 

 DNA 



H Potentially unaltered with no record of stocking or Suspected Unaltered 



contaminating species present 



J Suspected hybridized with records of contaminating species Potentially Altered 

 being stocked or occurring in stream 



N Hybridized and Pure populations co-exist in stream (use only Mixed Stock; Altered and 

 if reproductive isolation is suspected and testing completed) Unaltered co-exist 



Genetic tests can detect introgression between YCT and potentially introgressing species or 

 subspecies by finding alleles unique ("diagnostic alleles") to that potentially introgressing 

 species or subspecies within YCT populations. The number, and thus the proportion, of 

 potentially introgressing species or subspecies "diagnostic" alleles within YCT populations, is 

 used to estimate the level of introgression. One consequence of this approach is that proving a 

 stock of YCT to be genetically pure is essentially impossible: all individuals in a population 

 would have to be tested. Therefore, sample size must be considered when evaluating the 

 reliability of any genetic test. Generally, sample sizes should be large enough to determine, with 

 a pre-determined level of statistical reliability (95% has often been used), that a 1% or less level 

 of introgression would be detected. Both the number offish sampled and the number of alleles 

 that are "diagnostic" between species or subspecies determine the sample size needed for a pre- 

 determined level of statistical reliability. Thus, when genetic testing finds no evidence of 

 introgression, sample size is very important for assessing how valid the result may be. For this 

 assessment we reported results of all genetic testing, regardless of sample size, and then 

 displayed and summarized sample sizes for all genetic testing. 



Different genetics laboratories, and somefimes even the same lab, may report genetic results 

 differently; consequently, it can be difficult to compare genetic results across broad geographic 

 areas. Especially when only brief summaries of these data are stored in standardized fish 

 resource databases. An example of where this type of problem may occur is that of a mixed 

 stock population, where some individuals within the population may be genetically unaltered 

 YCT and other individuals may be genetically unaltered rainbow trout (RBT). Unless either the 

 local fisheries professional or the database indicated that non-random mafing was occurring 

 (code N; Table 1 ), we assumed genetic results were a fianction of random mating. If random 



Page - 7 



