Correlations Between Heterozygosity and Reproductive Success in the Blue Tit (Cyanistes Caeruleus): An Analysis of Inbreeding and Single Locus Effects

Abstract

To understand the mechanisms behind heterozygosity-fitness correlations (HFC), it is necessary to employ large numbers of markerswith known function and independently estimate the variation in inbreeding in the population. Here we genotyped 794 blue tits with 79 microsatellites that were distributed across 25 chromosomes and that were classified either as "functional" (N = 58) or "neutral" (N = 21). We found a positive effect of individual heterozygosity at multiple loci on clutch size, on the number of eggs sired by males, and on the number of recruits produced by males and females. We documented the occurrence of some consanguineous matings and found evidence for a particular type of population structure that can contribute to the occurrence of inbreeding. As the set of "neutral" loci provided more power to detect HFC and identity disequilibrium, we argue that "neutral" markers are better predictors of the effects of inbreeding. The number of significant effects at single loci did not exceed the expected number of false positives and no strong effects were associated with heterozygosity at "functional" markers. Thus, the HFC found here cannot be attributed to strong effects of the loci under study. K E Y W O R D S :EST-SSR, heterozygosity-fitness correlation, identity disequilibrium, microsatellites, population structure.

The strength of the association between heterozygosity and probability of interannual local recruitment increases with environmental harshness in blue tits

“…Previous studies carried out on this species found significant heterozygosity–fitness correlations considering different life‐history traits (reproductive performance: Foerster, Delhey, Johnsen, Lifjeld, & Kempenaers, 2003; García‐Navas, Ortego, & Sanz, 2009; Olano‐Marin, Mueller, & Kempenaers, 2011b; survival probability: Olano‐Marin, Mueller, & Kempenaers, 2011a; and parasite resistance: Ferrer, García‐Navas, Sanz, & Ortego, 2014). Consanguineous matings have been reported in these populations (Ferrer et al., 2014; García‐Navas et al., 2009), which may increase variance in inbreeding and the chance of detecting HFC due to extensive ID across the genome (Balloux et al., 2004; Szulkin et al., 2010).…”

“…Following the approach detailed in the study of Olano‐Marin, Mueller, and Kempenaers (2011a), Olano‐Marin, Mueller, and Kempenaers (2011b), we classified our markers as presumably functional or neutral by considering whether the genomic region where they are located is transcribed to RNA (Table S1) (Ferrer et al., 2014; Olano‐Marin, Mueller, & Kempenaers, 2011a; Olano‐Marin, Mueller, & Kempenaers, 2011b). More precisely, loci that were designed based on or showed homology to zebra finch ( Taeniopygia guttata ).…”

“…On the one hand, we refer to presumably functional microsatellite loci as those located within genomic regions that are transcribed to RNA, but we have no information on whether they are functional themselves (sensu Li et al., 2004). On the other hand, we did not aim to employ a “candidate gene approach” (e.g., to link fitness with specific genes/alleles with known function) or target microsatellite markers linked to a specific gene, but just to compare our two subsets of loci classified according with the above‐described criteria (e.g., Ferrer, García‐Navas, Bueno‐Enciso, Sanz, & Ortego, 2015; Ferrer et al., 2014; Olano‐Marin, Mueller, & Kempenaers, 2011a; Olano‐Marin, Mueller, & Kempenaers, 2011b). DNA extraction, microsatellite amplification, and genotyping and tests for deviations from Hardy–Weinberg equilibrium (HWE) and linkage disequilibrium (LD) between each pair of loci were performed as described in Ferrer et al.…”

“…Effect size was calculated for each locus as the partial correlation coefficient obtained from its respective model (Nakagawa & Cuthill, 2007). Finally, we used a general linear model (GLM) to analyze whether absolute effect sizes of single‐locus heterozygosities were associated with marker variability (allelic richness and observed heterozygosity, included as covariates) and differed between neutral and putatively functional loci (marker category was included as a fixed factor) (e.g., Ferrer et al., 2014; Olano‐Marin, Mueller, & Kempenaers, 2011a; Olano‐Marin, Mueller, & Kempenaers, 2011b). …”

“…Further, these microsatellite markers were classified as putatively neutral (14 loci) or functional (12 loci) by considering whether the genomic region where they are located is transcribed to RNA (sensu Olano‐Marin, Mueller, & Kempenaers, 2011a; Olano‐Marin, Mueller, & Kempenaers, 2011b), which allowed us to test differences in HFC and genetic diversity–environment interactions between these subsets of loci (see Methods for more details on the employed loci and their classification). Neutral markers can produce HFC either by general effects or by local effects if they happen to be linked to functional loci, whereas direct effects and strong local effects are only likely to be caused by functional markers (Ferrer et al., 2014; Olano‐Marin, Mueller, & Kempenaers, 2011a; Olano‐Marin, Mueller, & Kempenaers, 2011b). First, we analyzed (1) the relationship between probability of interannual local recruitment and individual genetic diversity and tested whether such associations (2) varied between putatively functional and neutral loci and (3) were better explained by a genomewide effect (“general effect hypothesis”) or strong linkage disequilibrium between the employed markers and genes under selection (“local effect hypothesis”).…”