Table 2.

Competing models of SNP evolution. Models 1–4 allow adaptation in a single population. Models 5–7 allow various forms of parallel adaptation to treatments. Each of the k allele frequency parameters were estimated with maximum likelihood

Model	Process	Model Parameters	k
0	Random drift	p 0: all populations	1
1	A1 adaptation	p 0: A2, B1, B2	2
		p 1: A1
2	A2 adaptation	p 0: A1, B1, B2	2
		p 1: A2
3	B1 adaptation	p 0: A1, A2, B2	2
		p 1: B1
4	B2 adaptation	p 0: A1, A2, B1	2
		p 1: B2
5	Parallel adaptation	p 0: A1, A2	2
		p 1: B1, B2
6	No Bee parallelism	p 0: A1, A2	3
		p 1: B1
		p 2: B2
7	Bee parallelism	p 0: B1, B2	3
		p 1: A1
		p 2: A2
8	Independence	p 0: A1	4
		p 1: A2
		p 2: B1
		p 3: B2