Table 2. African isolates of C. neoformans var. grubii are more diverse than the global population: comparison of the diversity indices and neutrality tests at each locus.

Locus	Populationa	No. of isolatesb	Gene diversity	Nucleotide diversity	Mean pairwise difference	Tajima's
			(h)	(π)	(d)	TD c
GPD1	African	99	0.86	0.005	2.3	−0.67
	Global	57	0.35	0.001	0.51	−1.8
TEF1	African	99	0.84	0.003	2.01	1.07
	Global	57	0.72	0.002	1.29	−1.62
CAP59	African	97	0.6	0.002	0.97	−0.38
	Global	47	0.61	0.002	1.01	−1.6
PLB1	African	80	0.77	0.003	1.89	−1.12
	Global	57	0.68	0.002	0.94	−0.89
IGS1	African	93	0.81	0.01	7.02	−1.7
	Global	41	0.23	0.0008	0.57	−1.7
SOD1	African	94	0.6	0.009	5.17	0.4
	Global	57	0.03	0.0005	0.28	−2.1
LAC1	African	88	0.82	0.006	2.67	0.45
	Global	49	0.73	0.004	1.3	1.02
URA5	African	92	0.75	0.007	4.7	0.24
	Global	47	0.6	0.001	1	−1.59

^aThe diversity of clinical and environmental strains of C. neoformans var. grubii from Botswana and South Africa was compared with that of a selection of the most genetically diverse global isolates available [10]. The “African” population included all clinical and avian isolates of C. neoformans var. grubii (85 strains) as well as 14 isolates from African trees; that is, we analyzed only one isolate per tree because multiple isolates from the same tree and surrounding soil were clonal. The “Global” sample was comprised of a larger sample of 57 previously genotyped, non-arboreal strains from 14 different countries (excluding Botswana and South Africa), and they included multiple isolates of each unique MLST or AFLP genotype [10]. Except for the CAP59 locus, this purposely enlarged and intentionally diversified global sample revealed less variation (i.e., lower h, π and d values) than the sample of African isolates.

^bWe excluded strains with recombinant haplotypes at a locus.

^cValues of Tajima's T_D for which the null hypothesis of neutrality was rejected are bolded.