Computational protocol: Sexual dimorphism dominates divergent host plant use in stick insect trophic morphology

Similar protocols

Protocol publication

[…] Electron micrographs were subjected to geometric morphometric analyses using 11 homologous landmarks chosen to outline the features listed above. Left (hereafter lingual) and right (hereafter occlusial) mandibles were analysed separately. Standardized landmark configurations were imported into MorphoJ [], where a generalized least squares Procrustes superimposition (GPS) was performed. The GPS generated a consensus configuration for both mandibles and a thin-plate spline was used to quantify deviation of individual specimen landmark configurations from the consensus along principal axes of shape change [,]. Deviations from the consensus along principal axes generate partial warps quantifying individual based shape changes [,]. For both mandibles, the thin-plate spline produced 16 partial warps and two uniform components per individual, representing local and overall shape variables, respectively. All 18 shape variables were used in subsequent analyses. [...] To quantify the relative importance of genetic background (e.g., population of origin) and rearing environment on mandible shape(s), we used individuals from populations LA and VPC reared from first instar to sexual maturity on both hosts in a reciprocal transplant experiment performed in the laboratory (TableĀ ; []). Landmark configurations for these samples were isolated and used to generate a new consensus configuration and a new set of size corrected shape variable as described above. We then used MANOVA to determine the influence of sex, population of origin, host rearing condition and their interactions on mandible shape variation. The population effect tested influences of genetic background while host rearing condition tested environmental effects. Interaction terms tested sex specific responses in the different populations and hosts, and population specific responses to hosts [,]. MANOVA was performed using default settings and a type III sum of squares model taking into account uneven sample sizes in STATISTICA 64 v.10 (Stat Soft Inc., Tulsa, OK USA). […]

Pipeline specifications