Parental genetic similarity and offspring performance in blue tits in relation to brood size manipulation

Abstract In birds, as in many other taxa, higher genetic similarity of mates has long been known to reduce offspring fitness. To date, the majority of avian studies have focused on examination whether the genetic similarity of social mates predicts hatching success. Yet, increased genetic similarity of mates may also reduce offspring fitness during later life stages, including the nestling period and beyond. Here, we investigated whether parental genetic similarity influences offspring performance using data from free‐living blue tits (Cyanistes caeruleus) collected across three breeding seasons. Additionally, we tested whether brood size manipulation affects the magnitude and direction of the relationship between genetic similarity of mates and offspring performance. Sixteen microsatellite markers were used to measure genetic similarity between biological parents. We found that the genetic similarity of parents negatively affects offspring immune response and this effect was independent of the experimental brood size manipulation.

The available evidence, therefore, suggests that the relationship between genetic similarity and offspring fitness is detected and/or more pronounced under harsher environmental conditions and may be nonsignificant under favorable conditions. Indeed, Keller, Grant, Grant, and Petren (2002) found negative effects of mating between close relatives in Darwin's finches only under nutritional constraints.
However, to our knowledge no experimental research has addressed this question.
Here, we investigate the relationships between genetic similarity of parents and offspring performance in the blue tit (Cyanistes caeruleus). Specifically, we investigated whether genetic similarity within the pair influences offspring quality in terms of body weight, tarsus length, and immunocompetence. We expected to see negative effects of parental genetic similarity on these three measured traits.
If the relationship between genetic similarity of mates and offspring quality and its magnitude depend on environmental conditions, one can expect a significant interaction between parental genetic similarity and environmental conditions. We therefore manipulated the brood size to alter conditions of nestling growth, and we expected to see a negative relationship between genetic similarity of mates and nestling quality that will be particularly pronounced among offspring from experimentally enlarged broods.

| Study site and experimental procedure
Our study was carried out on Gotland, Sweden (57°03′N, 18°17′E), during April-June, from 2009 to 2011. Adult birds were caught while feeding nestlings (May-June), using nest-box traps or mist nets. Birds were bled by brachial venepuncture for genetic analyses.
All nestlings were weighed on days 2 and 14 prior to the blood sampling and on days 11 through 12 as a part of the immune response assay. Additionally, on day 14 after hatching, nestlings were measured for tarsus length and ringed. To assess nestling T-cell-mediated immune response, on day 11 posthatching nestlings were injected with a nonpathogenic antigen, phytohemagglutinin (PHA) into their right wing web, and 24 hr later, the thickness of the wing web was measured with a pressure-sensitive calliper (see details in Drobniak et al. (2010) and Arct, Drobniak, Podmokła, Gustafson, and Cichoń (2013), Arct et al. (2017)). The measurements were taken by a single person and were highly repeatable (r = 0.92, F 461,924 = 36.1, p < 0.0001). The mean value of the three repeats was used in further analyses.
Here, we analyzed data on blue tits subjected to a brood size manipulation treatment, which is known to alter conditions of nestling growth and has a negative effect on various nestling characteristics (Cichoń & Dubiec, 2005;Neuenschwander, Brinkhof, Kolliker, & Richner, 2003). The following treatment description takes into account the cross-fostering experiment that was performed on all nests by exchanging halves of broods between nests and performed in another study (for details see Drobniak et al., 2010). Half of the nestlings were exchanged between control and experimental nests. Briefly, on the second day after hatching, we matched pairs of broods according to the same hatching date and similar brood size (±1 chick) (in
TA B L E 1 Linear mixed model analyses (LMMs) of the body mass on day 14 (g), tarsus length (mm), and T-cell-mediated immune response to PHA Nest of rearing, nest of origin, and the year of study were included as higher-level random effects. Parental genetic similarity was entered as a covariate; experimental treatment (experimental nests/control nests) and offspring sex (female/male) were defined as fixed factors. In the analysis of the Tcell-mediated immune response, we used the body mass on day 12 (when the immune response was measured) as a covariate. We present two types of R 2 -marginal R 2 (m) and conditional R 2 (c) for both LMMs. Bold indicates significant effects (p > 0.05).

| Statistical analysis
We used a linear mixed model (LMM) in R (R Development Core  (Table S1), which allow us to interpreted main results (e.g., Schielzeth, 2010).
In the analysis of the T-cell-mediated immune response, we used the body mass on day 12 (when the immune response was measured) as a covariate. The data met the assumptions of a linear mixed model (which we judged visually using model residuals). We used the "MuMIn" package in R to estimate the coefficient R 2 from the mixed models (Bartoń, 2009;Nakagawa & Schielzeth, 2013).  (Table S2). We confirm that the total number of excluded observations and the reasons for making these exclusions have been reported in the Method section.

| RE SULTS
The interaction between parental genetic similarity and experimental treatment appeared nonsignificant for all measured traits (Table   S1) and was thus deleted from the final model.
We found a significant negative correlation between T-cell-mediated immune response to PHA and parental genetic similarity (Table 1, Figure 1). There were no effects of parental genetic similarity on body mass and tarsus length on day 14. Experimental treatment significantly affected the body mass of nestlings (means ± SD: control nests 11.25 ± 0.86 g; experimental nests 10.84 ± 1.14 g (Table 1) but had no significant effect on PHA immune response and tarsus length.

| D ISCUSS I ON
In this study, we investigated the effect of brood size manipulation treatment on the relationship between genetic similarity and offspring performance. To our knowledge, the effect of environmental condition on the relationship between parental genetic similarity and offspring performance has never been experimentally investigated.
However, we failed to find any evidence that the brood size manipulation experiment affects the relationship between genetic similarity and offspring performance. There is accumulating evidence that brood size manipulation has an effect on nestling immunocompetence and body condition (Horak, Tegelmann, Ots, & Moller, 1999;Sanz & Tinbergen, 1999). It is probable that brood enlargement creates a competitive environment for the offspring. Here, we showed that enlargement of the brood size affected nestling body mass on day 14. However, in the case of PHA immune response and tarsus length, we observed only a trend that offspring from experimental nests had lower immunocompetence and shorter tarsi. Such discrepancy may be a result of a relatively small sample size. This obviously reduces the power to find a significant interaction between experimental treatment and parental genetic similarity. Our previous studies on blue tits also did not support the idea that heterozygosity-fitness correlations become stronger under stressful conditions than under optimal conditions (Arct et al., 2017). This does not, however, exclude the possibility that other environmental factors, such as parasite prevalence, food availability, or harsh weather conditions, might strengthen the association between parental genetic similarity and offspring fitness.
Here, we showed that genetic similarity between pair members had a significant effect on offspring immunocompetence in blue tits. Specifically, in line with our prediction, we found a negative Genetic similarity between parents PHA immune response relationship between parental genetic similarity and offspring PHA immune response (Figure 1). Similar to the previous study on great tits (P. major) (van de Casteele et al., 2003), we showed that the effect of parental genetic similarity on offspring fitness-related traits is not limited to embryonic stage but may also be important at later life stages. The ability to mount a strong immune response to pathogens was found to play an important role in determining individual survival prospects (Cichoń & Dubiec, 2005). Indeed, immune response was positively related to the probability of survival in nestlings (Cichoń & Dubiec, 2005) as well as adult birds (Gonzalez et al., 1999;Saino, Calza, & Møller, 1997), and the genetic variability in immune response to pathogens in natural populations is well documented (Cichoń, Sendecka, & Gustafsson, 2006;Drobniak et al., 2010;Saino et al., 1997 In our study, the negative relationship between parental genetic similarity and offspring fitness-related traits was visible only in the immune response. Similarly, in our previous study on blue tits we found that the immune response of extra-pair young to phytohemagglutinin was stronger than that of within-pair half-sibs, but at the same time the superior quality of EPY was not confirmed in terms of body mass and tarsus length. This could theoretically be due to a causal relationship between the individual level of genetic diversity and the PHA immune response (Fossøy, Johnsen, & Lifjeld, 2008). However, a correlation between the PHA immune response and individual level of heterozygosity was not supported in our previous studies on the same population of blue tits (Arct et al., 2017). In contrast, we found a positive relationship between individual heterozygosity and body mass of female nestlings 14 days posthatching. Thus, the observed relationship between parental genetic similarity and offspring immunocompetence may potentially be explained not only through genetic effects but also by maternal effects. Indeed, there is evidence that females may adjust their reproductive investment in response to the genetic similarity of their partners (Arct, Rutkowska, Martyka, Drobniak, & Cichoń, 2010). For example, differences in offspring performance may result from nongenetic maternal effects (Martyka, Rutkowska, & Cichoń, 2011

ACK N OWLED G M ENT
We thank Edyta Podmokła for conducting the parentage genetic analyses. This research was funded from resources of the National Science Centre, allocated on the basis of the decision number DEC2013/09/B/ NZ8/03322 (OPUS 9) and UMO-2015/18/E/NZ8/00505.

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R CO NTR I B UTI O N S
AA-contributed to the design of the work, genetic and statistical analysis, and interpretation of data, and wrote the manuscript; SMD-verified the analytical methods, designed the figures, and interpreted the data; SM-involved in genetic analysis; LG-supervised the long-term study and worked on the manuscript; MC-contributed to the design and implementation of the research, to the analysis of the results and to the writing of the manuscript.

DATA ACCE SS I B I LIT Y
All data and microsatellite genotypes are available through the Dryad Digital Repository: https ://doi.org/10.5061/dryad.v6r0758.