Fig 1.
The potential inputs that could affect trait values in phenotypic traits in merpeople.
Boxed parameters are the true parameters included in at least one of the traits. Unboxed parameters are included in models as potential sources of variance, but did not affect the simulated traits. a represents additive genetic effects, and Ma the maternal genetic effect. n is the total direct environmental effect, and Mn the maternal environmental effect. epi and Mepi are direct and maternal epigenetic effects, and c and Mc are direct and maternal social network/cultural effects. Paternal effects were not included in the simulations or models, but are shown here as a potential effect on trait values. F0 and F1 indicate generations.
Fig 2.
Reduced versions of the matrix information included in the animal models shown in this paper.
(a) The genetic relatedness matrix, derived from a pedigree, which contains the probabilities that individuals i and j share alleles that are identical by descent. (b) The environmental similarity matrix, which contains elements Sn(i,j) (written in the figure as S, as the same equation is used for elements of Sepi) measuring the Euclidean distance between the environment individuals experience, p and q being the scaled environmental vectors for individuals i and j across l environments. (c) Epiallelic similarity, which contains elements Sepi(i,j), measuring the Euclidean distance between individuals’ epigenetic profiles, where p and q are the scaled epigenetic vectors for individuals i and j, across l epigenetic islands. (d) The geodesic distance matrix (social network connectedness) that shows the proximity of individuals in the social network, based on minimum path lengths (gi,j).
Fig 3.
Variance components of tail-fin colour (trait 1).
This trait was simulated to contain additive genetic, environmental, and residual variance, each with mean = 0 and variance = 1. For each of the models, the pair of plots show the absolute values (left plot) and the proportion (right plot) of total trait variance explained. a) Model containing only additive genetic (a) and residual variance (r). b) Model including additive genetic (a), environmental (n), and residual variance (r) (the true model). c) Model including additive genetic (a), epigenetic (epi), and residual variance (r). d) Model containing additive genetic (a), social (c), and residual variance (r). The vertical dashed line shows simulated value for variance components (1).
Fig 4.
Variance components of body size (trait 2).
This trait was simulated to contain additive genetic, environmental, maternal environemental, and residual variance, each with mean = 0 and variance = 1. For each of the models, the pair of plots show the absolute values (left plot) and the proportion (right plot) of total trait variance explained. a) Model containing only additive genetic (a) and residual variance (r). b) Model including additive genetic (a), environmental (n), and residual variance (r). c) Model including additive genetic (a), environmental (n), and residual variance (r), and maternal variance estimated by including maternal identity as a random effect (M). d) Model including additive genetic (a), environmental (n), maternal environmental (Mn), and residual variance (r) (the true model). e) Model including additive genetic (a), epigenetic (epi), maternal epigenetic (Mepi), and residual variance (r). f) Model containing additive genetic (a), social (c), maternal social (Mc), and residual variance (r).
Fig 5.
Variance components of swimming speed (trait 3).
This trait was simulated to contain additive genetic, maternal genetic, environmental, maternal environmental, and residual variance, each with mean = 0 and variance = 1. For each of the models, the pair of plots show the absolute values (left plot) and the proportion (right plot) of total trait variance explained. a) Model containing only additive genetic (a) and residual variance (r). b) Model including additive genetic (a), environmental (n), and residual variance (r), and maternal variance estimated by including maternal identity as a random effect (M). c) Model including additive genetic (a), maternal genetic (Ma), environmental (n), maternal environmental (Mn), and residual variance (r)(the true model).