Table 4

Results of the Type one error analysis
Multilocus-Genotype (MLG)
 Math(g)
 ω (g)
 Ar(g) for C2N = 0.05
 C2Nfor Ar(g) = 0,0275
Ror1
 1.2196· 10-5
 0.232
 0.0275
 /
Ror2
 1.2658· 10-7
 0.054
 /
 0.0032
Ror3
 1.5866· 10-5
 0.090
 /
 0.0087
Ror4
 3.4244· 10-5
 0.268
 /
 0.0649
Wib1
 3.5754· 10-5
 0.282
 0.0162
 /
Wib2
 1.6788· 10-7
 0.071
 /
 0.0104
Wib3
 8.0274· 10-6
 0.282
 0.0162
 /
Wib4
 1.0481· 10-5
 0.166
 0.0275
 /
Wib5
 1.9321· 10-6
 0.038
 /
 0.0031
Wib6
 5.1381· 10-5
 0.282
 0.0161
 /
Wib7
 1.9983· 10-6
 0.224
 /
 0.0830
Wib8
 6.2936· 10-7
 0.052
 /
 0.0057
Wib9
 7.4621· 10-7
 0.083
 /
 0.0139
Wib10
 4.7596· 10-6
 0.282
 /
 0.1231
Wib11
 1.0203· 10-8
 0.038
 /
 0.0031
Wib12
 5.4225· 10-8
 0.064
 /
 0.0085
Wib13
 2.2618· 10-7
 0.038
 /
 0.0031

(1) the expected frequency Math(g) of the multilocus genotypes (MLG) in Table 3 under the assumption of random association of the alleles; (2) the least upper bound ω (g) of the frequency of the MLG set by the population frequencies of the alleles that are represented in this MLG; (3) the value of Ar(g) for which C2N = ε = 0.05 (/denotes the absence of a Type one error); (4) the value of C2N for the relative gene association Ar(g) = 0.0275 that would have avoided all of the observed Type one errors (values in bold exceed significance level ε = 0.05)

Höltken and Gregorius BMC Ecology 2006 6:13   doi:10.1186/1472-6785-6-13