Table 2. Bayesian testing of monophyly: posterior probability of monophyly for each dataset.

Group	COI + ND2 (mtDNA)	COI + ND2 (mtDNA)	YcfW (Wigglesworthia)	Period	ITS1
N individuals	29 (all haplotypes)	16	14	16	14
Nst (rates)a	6	6	6	2	1
G. f. fuscipes	<2.63×10−5	<2.63×10−5	<2.63×10−5	<2.63×10−5	0.917
G. f. martinii	0.136	0.363	0.085	0.999	NDd
G. f. quanzensis	<2.63×10−5	0.207	0.082c	4.74×10−4	0.130
Ethiopia	1.000	0.998	0.999	0.994	ND
LVB + martinii b	0.995	0.995	0.864	<2.63×10−5	ND

Significant test results are shown in bold font. For the mtDNA data set the 16 individuals that were sequenced at all loci were analysed together with and then separately from the remaining specimens for which only mtDNA data were available. This allowed comparison of similar sized data sets for mitochondrial, symbiont and nuclear DNA data sets. The discrepancy between the sample size for ITS1 and Wigglesworthia YcfW loci (n = 14) and the Period and COI+ND2 loci (n = 16) is due to the lack of Bena Tschibangu and Buvuma genotypes from the former.

^aPermitted number of nucleotide substitution rates.

^bA combination of flies from Lake Victoria Basin (LVB) and G.f. martinii.

^cResult is only for quanzensis samples from western DRC.

^dnot done.