Likelihood-based Species Identification Using Cytochrome B (Cytb) Gene As Exampled From Rat Snakes(Elaphe)

Wildlife Crime,including poaching,illegal obtaining,possessing and using,is one of the most important factors threatening global biodiversity.Specie identification is a critial step for law enforcement.Mt DNA is a widely used molecular genetic marker for species identification.Such approaches normally achieve conclusions through comparisons between evidence sequences and a few representative individuals of target species by either calculated identity or phylogenetic tree structure.Conclusions derived from such statistical methods can not be supportetd by quantified risk or relibility,and particularly risky for closely related species.The present study proposed a likelihood-based statistical method using cytochrome b gene of rat snakes as example,including Elaphe carinata(n=52),E.dione(n=15),E.quadrivirgata,(n=40),E.quatuorlineata(n=43),E.bairdi(n=22),and E.schrenckii(n=18).Frequency of each nucletide was calculated on a 43 bp fragement and a 75 bp fragement using population sequence data,based on which the assignment probability of given con-specific and hetero-specific samples could be calculated.The con-specific and hetero-specific assignment probability data were –lg transformed and subjected to construct a normal probability density function.The function was then used to estimate the probability of a given sample being assigned to a given species.Alternatively,con-specific and hetero-specific assignment proibability data were also used to estimate 99% inclusive confidence interval and exclusive confidence interval that could be used as the criteria for species inclusion and exclusion.Results demonstrated a significant partition between con-specific and hetero-specific probability distribution on both the 43 bp fragement and 75 bp fragement.Species identificaiton could be performed precisely using the probability density function and inclusive and exclusive confidence intervals.Comparing to the present statistical methods,this methods better describes the species characteristics by fully considering the weight of each nucleotide,expecially expressing the role of nucleotides with great diagnostic significance.Conclusions based on this method could be supported with quantified reliability or risk,better fulfilling the requirement of forensic identification.Meanwhile,high relibility was generated on fragements as short as 43 bp and 75 bp,allowing very short amplification for sequencing experiments on degraded evidence samples implying greater application potential.