Table 2.

Correspondence between the Main Results in This Article and in Edge and Rosenberg (2015)

Result	Equation Number
	This Article	Edge and Rosenberg (2015)
E(DT) = 0 due to symmetry around 0 of the distribution of DT.	35	36
$\text{Var}\left(D_T\right)=E\left(D_T^2\right)$ does not increase without bound with the number of loci and is equal to $D_L^2$, where DL is an analogue of DT for the allelic count at a single locus.	36, 37	37, 38
FST ≈ QST.	45, 47, 48	42, 43
As the number of loci k increases without bound, the genetic misclassification rate approaches 0.	55	5
The expectation of the approximate trait-based misclassification rate is closely related to the genetic misclassification rate obtained using one locus.	57	47

The main results in this article reduce to the main results in Edge and Rosenberg (2015) under the following assumptions: (a) The allele frequencies are the same at each locus, meaning that $p_i=\overline{p}=p$ for all i, $q_i=\overline{q}=q$ for all i, ${\delta }_i=\overline{\delta }=q-p$ for all i, and $s_p^2=s_q^2=s_{\delta }^2=0$. (b) The allele frequencies are symmetric, meaning that $\overline{q}=1-\overline{p}$. In conjunction with assumption (a), (b) implies that $F_{ST}={\overline{\delta }}^2=1-4pq$. (c) The organisms are haploid, or in the present article’s notation, ℓ = 1. Assumption (c) implies that $Q_{ST}={\rho }_T^2$ (Eqs. 26–28).