Table 1.
Ranked MuHiSSE results for body size and ecotype, in descending order of support for tested models.
Fig 1.
Time-calibrated phylogenetic tree of the Pomacentridae.
Lower half of the tree including the subfamilies Microspathodontinae and Chrominae. Time axis in millions of years before present. Nodal values with Bayesian posterior support levels above 0.9 are indicated with blue dots (see S2 Fig in S1 File for all support values and S3 Fig in S1 File for all ages). Representative photos are labeled with the first letter of genus and species, matching species names in the tree closest to the photo.
Fig 2.
Time-calibrated phylogenetic tree of the Pomacentridae.
Upper (crown) half of the tree including the subfamilies Glyphysodontinae and Pomacentrinae. Time axis in millions of years before present. Nodal values with Bayesian posterior support levels above 0.9 are indicated with blue dots (see S2 Fig in S1 File for all support values and S3 Fig in S1 File for all ages). Representative photos are labeled with the first letter of genus and species, matching species names in the tree closest to the photo.
Fig 3.
Ancestral state reconstruction of damselfish body size.
Maximum body size (total length) gap coded into 3 states with the discretized trait mapped onto the damselfish time-calibrated phylogeny. Transition rates between character states and their rate (if non-zero) between body size states during species diversification throughout the history of the Pomacentridae are shown in lower right. S = small (4.5–10.2 cm TL), M = medium (10.5–14.1 cm TL), L = large (14.4–45 cm TL). Reconstructed state transition counts are S -> M = 46, M -> S = 37, M -> L = 44, L -> M = 15, S -> L = 0, L -> S = 0.
Fig 4.
Ancestral state reconstruction of damselfish dietary ecotype.
Dietary ecotype traits of benthic, pelagic and intermediate foraging behavior mapped onto the damselfish time-calibrated phylogeny. Transition rates between character states and their rate (if non-zero) between ecotype states during species diversification throughout the history of the Pomacentridae are shown in lower right. Reconstructed state transition counts are: B -> I = 6, I -> B = 25, I -> P = 35, P -> I = 26, B -> P = 0, P -> B = 0.
Fig 5.
Mirror tree illustration of dietary ecotype and farming behavior.
The patterns of dietary ecotype history (left) are illustrated with the presence and absence of the behavioral trait of algal patch “farming” (right) on the damselfish time-calibrated phylogeny.
Fig 6.
Body size diversification rates on the time-calibrated phylogeny of the Pomacentridae.
Body size shows multiple areas of elevated diversification in lighter colors throughout the tree (Amblyglyphidodon, Amphiprion, Chromis, Plectroglyphidodon and Stegastes).
Fig 7.
Ecotype diversification rates on the time-calibrated phylogeny of the Pomacentridae.
The relatively low, steady pattern of diversification in the dietary ecotype trait is punctuated with several independent origins of elevated ecotypic diversification rate (Amphiprion is the most prominent) in lighter colors throughout the tree.
Fig 8.
The distribution of net diversification across (A) damselfish size classes, and (B) damselfish ecotypes. Model averaged mean and standard deviation of diversification for each size class and ecotype is indicated by a colored circle and bar, respectively. The height of the peak indicates the number of tips for each state with that particular estimated net diversification rate. Mean estimated net diversification was highest in the medium size class and the pelagic ecotype state.