Computational protocol: Clock gene variation in Tachycineta swallows

Similar protocols

Protocol publication

[…] Observed and expected heterozygosities for the Clock poly-Q region for each species were calculated using ARLEQUIN version 3.11 (). We tested for departures from Hardy–Weinberg equilibrium (HWE) using GENEPOP version 4 (; ) with parameters of 10,000 dememorization, 10,000 batches, and 10,000 iterations. We used contingency tables to test whether allele frequencies were different between years for each species.To test for a relationship between length polymorphism in the Clock poly-Q region and breeding phenology we used either mean Clock poly-Q allele size or the Clock poly-Q genotype. We examined the effect of Clock poly-Q region on lay date in separate analyses for each species using linear mixed models in which lay date was the dependent variable, female identity (band number) as a random effect to account for repeated measure for same females in different years, and breeding year and mean Clock poly-Q allele size (or Clock poly-Q genotype) as fixed effects. The effect of Clock poly-Q region on incubation duration was examined using linear mixed models in which incubation duration was the dependent variable, female's band number as a random effect and year, lay date, clutch size, and mean Clock poly-Q allele size (or Clock poly-Q genotype) as fixed effects. For T. bicolor, we also had information on female's age group (classified as “first year” or “after first year”) thus female's age was included as a fixed effect in all T. bicolor models. Data on incubation duration were not available for T. meyeni, therefore we were not able to examine the relationship between Clock poly-Q region and incubation duration for this species.To examine fitness consequence on timing of breeding, we used generalized linear mixed models (SAS 9.1, SAS Institute Inc., Cary, NC) for the relationship between clutch size and lay date, with female's band number as a random effect and year and lay date as fixed effects (female's age was included as fixed effect in T. bicolor model). The effects (slopes) generated from this model of seasonal decline in clutch size were also used to examine the relationship between number of Clock poly-Q alleles and the effect of seasonal decline in clutch size using Spearman rank-order correlation. We used this test also to examine the correlation between population Clock poly-Q mean allele size and latitude. […]

Pipeline specifications

Software tools Arlequin, Genepop
Application Population genetic analysis