The reversibility of phenotypic evolution is likely to be strongly influenced by the ability of underlying developmental systems to generate ancestral traits. However, few studies have quantitatively linked these developmental dynamics to traits which re-evolve. In this study we assess how changes in the inhibitory cascade, a developmental system that regulates relative tooth size in mammals, influenced the loss and reversals of the posthypocristid, a molar tooth crest, in the kangaroo superfamily Macropodoidea. We find that posthypocristid loss is linked with reduced levels of posterior molar inhibition, potentially driven by selection for lophodont, higher-crowned molar teeth. There is strong support for two posthypocristid reversals, each occurring after more than 15 million years of absence, in large-bodied species of Macropus, and two giant extinct species of short-faced sthenurine kangaroo (Procoptodon). We find that whereas primitive posthypocristid expression is linked to higher levels of posterior molar inhibition, re-emergence is tied to a relative increase in third molar size associated with increasing body mass, producing molar phenotypes similar to those in mouse where the ectodysplasin pathway is up-regulated. We argue that although shifts in the inhibitory cascade may enable re-emergence, dietary ecology may limit the frequency of phylogenetic reversal.
crownH_data
Extant macropodoid crown height data based on measurements of museum specimens. "ID" specimen collection number,"Side" = describes if the specimen is a left (L) or right (R) dentary."Source" = collection holding specimen (AM= Australian Museum Sydney, SAM= South Australian Museum Adelaide, WAM= Western Australian Museum Perth, FU= Flinders University Vertebrate Palaeontology Laboratory Adelaide , ANWC= Australian National Wildlife Collection Canberra, NTM= Northern Territory Museum, Alice Springs), "PW"= posterior width, "PHb"=Molar crown height measured at the hypoconid, "Tooth_identity"= molar position of the tooth.
FAD
Chronostratigraphic data on macropodoid first appearances in the fossil record. Data was compiled in Excel and exported as text file. As described in the papert the "FAD" corresponds to the earliest representative of the lineage while "LAD" refers to the first appearance of the tip taxon associated with that respective branch. All numeric values refer to millions of years before present (Myr). Tip taxa correspond to those in the morphological phylogeny.
Molecular_tree_states
Tree states relating to a pruned version of the molecular phylogeny originally produced by Mitchell et al. (2014; see paper text for full citation). Data was compiled in Excel and exported as text file. "1" = Posthypocrsitid present, "0" = Posthypocrsitid absent.
Morphological_tree_states
Character states for the modified morphological phylogeny originally produced by Prideaux and Warburton (2010; see in paper text for full citation). "1"= Posthypocrstid presemy, "0"= posthypocristid absent. Data was originally compiled in excel and exported as a text file.
Morphological phylogeny
Modified (see paper text) most parsimonious morphological input tree from Prideaux and Warburton (2010).
mpts.tre
Timescaled morphological phylogeny of Macropodoidea
Output timescaled morphological phylogeny from R in Newick tree file.
mpts_Tscale.tre
Posthypocrstid_relief_data
Posthypocristid relief data based on CT scan measurements in Avizo 8. PHCvert=vertical posthypocristid relief, PHCwdith= width across the talonid fossa at its point of maximum relief, Relief_Ratio = the quotient of PHCvert by PHCwidth. Comment= describes whether the crest is a homologue or analogue of the posthypocristid.
PHC_relief_data.txt
Tooth_dimensions
Tooth dimension data based on museum collection specimens."ID"= Specimen number, "Side"= specimen is left ("L") or right ("R") dentary, "Origin"= collection holding specimen (AM= Australian Museum Sydney, BMNH= British Museum of Natural History, SAM= South Australian Museum Adelaide, UCMP= University of California Museum of Paleontology, QVM= Queen Victoria Museum, Inveresk, Tasmania, WAM= Western Australian Museum Perth, FU= Flinders University Vertebrate Palaeontology Laboratory Adelaide , ANWC= Australian National Wildlife Collection Canberra, NTM= Northern Territory Museum, Alice Springs, QM=Queensland Museum, Brisbane), "AW"= maximum anterior molar width, "PW"= maximum posterior width, "L"= maximum molar length, "Tooth_id" = molar tooth position, "Collector"= observer responsible for measurement (AC= Aidan Couzens, GJP= Gavin J. Prideaux).
SurfaceFiles
Surface files exported from MeshLab v1.3.3 64bit. Naming convention is "Institution_SpecimenNumber_Species_Element.ply". Institutional abbreviations:
Australian Museum (AM), Australian National Wildlife collection (ANWC), Flinders University Vertebrate Palaeontology Laboratory (FU), Northern Territory Museum
Alice Springs (NTM), Queensland Museum (QM), South Australian Museum (SAM), University of California Musuem of Paleontology (UCMP), Wesetern Australian Museum (WAM). Element abbreviations:
LRMX= Lower right (or left;L) molar (X denotes tooth number (1-4); X if unknown), OES= Outer enamel surface, EDJ= Enamel-dentine junction, ENAM= Enamel cap, DENT= Dentine, RMAND= Right (or left;L) mandible.
Sum_m3HI_byW
Summary Crown heigh data for living macropodoids exported from R. Taxa= Species, MeanPhyb= Hypsodonty index calculated as mean hypoconid crown height divided by max talonid width, sePhyb= bootstrap estimate of standard error.
Posterior_sequence_ratios
Summary posterior molar proprtion data (m2-4) for Macropodoid taxa exported from R. Summary statistics (mean, sample size; n, standard deviation;sd) for m3/m2 and m4/m2.
R_scripts
R script files compiled in the program TinnR.