Uncertainty in the mating strategy of honeybees causes bias and unreliability in the estimates of genetic parameters

Table 4 Errors in genetic parameter estimates for sire modeling scenarios relative to open mating

Sire modelling scenario	\({\widehat{\upsigma }}_{{\text{W}}}^{2}\)			\({\widehat{\upsigma }}_{{\text{Q}}}^{2}\)			\({\widehat{{\text{r}}}}_{{\text{WQ}}}\)
Sire modelling scenario	Relative bias (%)	Relative SE (%)	% strong deviations	Relative bias (%)	Relative SE (%)	% strong deviations	Bias	SE
O_NoPheno	− 3.02	23.09	40	1.04	21.74	34	− 0.021	0.195
O_TwoPS_P	64.45	22.34	98	21.03	20.75	54	− 0.245	0.110
O_FixedGroup	− 3.04	19.39	34	0.83	18.37	24	0.003	0.146
O_RandGroup	− 2.49	18.91	33	1.24	17.84	23	− 0.003	0.137

Sire modelling scenarios: O_NoPheno: modeling a single PS_P and open-mated DPQ’s colony phenotypes excluded; O_TwoPS_P: modeling one open mating PS_P per open mating drone subpopulation; O_FixedGroup: modeling a single PS_P and drone subpopulations accounted for by a non-genetic fixed effect; O_RandGroup: modeling a single PS_P and drone subpopulations accounted for by a non-genetic random effect
\({\widehat{\upsigma }}_{{\text{W}}}^{2}\), \({\widehat{\upsigma }}_{{\text{Q}}}^{2}\) and \({\widehat{{\text{r}}}}_{{\text{WQ}}}\): estimates of the genetic variances and correlation for worker and queen effect
Strong deviations are those that differ by more than 20% from the true values. The relative bias was calculated as \(\left(\frac{1}{{n}_{rep}}\sum_{1}^{{n}_{rep}}\left(\frac{{\widehat{\upsigma }}_{{\text{A}}}^{2}-{\upsigma }_{{\text{A}}}^{2}}{{\upsigma }_{{\text{A}}}^{2}}\right)\right)\). The relative standard error (SE) was the standard deviation across replicates of the relative difference between estimates and the true value

ISSN: 1297-9686