|
Sources of variation
|
df
|
F
|
P
|
|---|
|
Log-transformed LPS-specific immune response (mOD min− 1); N = 261
|
|
Year
|
1, 40.70
|
91.51
|
< 0.001
|
|
Maternal immunization
|
1, 50.39
|
0.72
|
0.402
|
|
Brood size manipulation
|
1, 8.73
|
0.37
|
0.557
|
|
Nestling status
|
2, 82.04
|
3.26
|
0.043
|
|
Offspring sex
|
1, 222.74
|
3.39
|
0.067
|
|
Hatchling body mass
|
1, 234.87
|
2.88
|
0.091
|
|
Maternal immunization × brood size manipulation
|
1, 239.95
|
4.81
|
0.029
|
|
Maternal immunization × offspring sex
|
1, 221.43
|
3.11
|
0.079
|
|
Log-transformed total Ab production (mOD min−1); N = 257
|
|
Year
|
1, 42.67
|
21.40
|
< 0.001
|
|
Maternal immunization
|
1, 58.05
|
0.41
|
0.524
|
|
Brood size manipulation
|
1, 244.24
|
0.29
|
0.588
|
|
Nestling status
|
2, 82.66
|
0.04
|
0.961
|
|
Offspring sex
|
1, 228.09
|
1.22
|
0.270
|
|
Hatchling body mass
|
1, 238.20
|
3.63
|
0.058
|
- Table 2 presents the results of linear mixed models that examined the effects of a set explanatory variables on log-transformed LPS-specific immune response and log-transformed total Ab production (both specific immune response and total Ab production were estimated as the differences between post- and pre-immunization Ab titres). Full models included year (to control for inter-season differences), maternal immunization (control vs. immunized females), brood size manipulation (control vs. enlarged broods), nestling status (to control for cross-fostering effects; there were three levels of the factor: nestlings from cross-fostered nests moved to foster broods, nestlings from cross-fostered nests that stayed in their original broods and nestlings from non-cross-fostered nests where all nestlings stayed in their original nests) and offspring sex (to control for differences between male and female nestlings) as fixed factors, and hatching date, clutch size, hatchling body mass (2 days after hatching), log-transformed LPS-specific Ab titres (only in the analysis of LPS-specific immune response) and log-transformed total Ab titres (only in the analysis of total antibody production) in 5-day-old nestlings as covariates. All two-way interaction terms between maternal immunization, brood size manipulation and offspring sex were tested as well. Nest of origin (female identity) and nest of rearing (foster female identity) were included in all models as random factors (results not shown). Presented are reduced models after sequential backward elimination of non-significant (if P ≥ 0.10) interactions and covariates
