https://doi.org/10.5061/dryad.h18931zsh

Authors:
Christopher M. Martinez (UC Irvine, co-first author)
Katherine A. Corn (UC Davis, Virginia Tech, Washington State University, co-first author)
Sarah Williamson (UC Davis)
Darien Satterfield (UC Davis)
Alexus S. Roberts-Hugghis (UC Davis)
Anthony Barley (Arizona State University–West Campus)
Samuel R. Borstein (University of Michigan)
Matthew D. McGee (Monash University)
Peter C. Wainwright (UC Davis)

Contact (Correspondence authors):
c.martinez@uci.edu (C.M.M); cornkatherineap@gmail.com (K.A.C)

Study Summary:
We examine functional and morphological diversification in 300 cichlid species from three African Great Lake radiations (Victoria, Malawi and Tanganyika) and a continental radiation in the Neotropics. We show that trait variation in Lake Victoria, the youngest radiation, was a subset of that within Lake Malawi, which itself was nested within the older Tanganyikan radiation. Functional diversity in Neotropical cichlids was often lower than in Lake Tanganyika, despite being much older. This work suggests that cichlid functional evolution has played out in strikingly similar fashion in different radiations, with contingencies eventually resulting in lineage-specific novelties.

Software:
Analyses were done in the R statistical environment, version 4.2.0

R Packages used:
ape, version 5.6-2
geiger, version 2.0.7
geomorph, version 4.0.3
ggplot2, version 3.4.4
hypervolume, version 2.0.12
MASS, version 7.3.58.2
phytools, version 1.0-3
plyr version 1.8.8
reshape2, version 1.4.3
sfsmisc, version 1.1.15
tidyverse, version 1.3.1

Description of the data and file structure

The information below describes the contents of data files and code associated with this study:

1. info_table.csv: table of length equal to number of motion shapes (11,100) with the following columns:
   a. specimen: label with genus, species and specimen identification number
   b. genus_species: species names for cichlids used in this study
   c. subfamily: cichlid subfamily that species comes from
   d. locality_riftlake: the radiations that the species comes from
   e. strike_number: denotes the 10 shapes associated with each motion
   f. phylo_species: names of species in phylogeny
2. Mcgee_2020_tree.tre: phylogenetic tree of cichlids from McGee et al. 2020.
3. phylo matching.csv: table to match species names in study to names in phylogeny
4. motion_lm_data.csv: landmark data for 1,110 motions analyzed in this study
5. Classifiers.txt: table of length equal to number of species (300) with the following columns:
   a. genus_species: species names for cichlids used in this study
   b. phylo_species: names of species in phylogeny
   c. region: radiation locality of cichlid species
6. Avg Kinematic Components (logged _ scaled).txt: species-averaged and scaled motion component traits, including premaxillary protrusion, maxillary rotation, lower jaw rotation, cranial elevation, hyoid depression, and gape.
7. Avg. START head shapes, ALIGNED.txt: species-averaged, Procrustes-aligned landmark head shape data
8. Avg. Trajectory Shapes, ALIGNED.txt: species-averaged, Procrustes-aligned motion pattern (i.e., motion trajectory) data.
9. kinesis traits.txt: species-averaged composite kinesis-derived traits, including kinesis, the logged value of kinesis skew, and the logged value of kinesis coefficient.
10. PCs 1-4, START head shapes.txt: First four principal component axes from a principal component analysis (PCA) on interspecific head shapes.
11. PCs 1-4, TRAJECTORY Shape, ALIGNED.txt: First four principal component axes from a principal component analysis (PCA) on motion pattern data.
12. PCs 1-6, COMPONENTS.txt: Principal component axes from a principal component analysis (PCA) on motion components.
13. sample videos.zip: Folder containing 12 videos in .avi file format analyzed in this study. Videos show a representative range of feeding kinematics. File names for videos are in format, Genus_species_specimen_video (e.g. Altolamprologus_calvus_02_017).
14. Sample frames, landmark data.zip: For each video from “sample videos.zip”, an associated folder with 10 still frames in .tif format and a .tps file with anatomical landmark data for those frames. The .tps files can be opened with the digitizing software tpsDig2 (https://sbmorphometrics.org/soft-dataacq.html) or imported into R with the readland.tps function in the R package geomorph.

Code/Software

1. cichlid data preparation, plots.R: code for preparing and plotting kinematic trait data.
2. cichlid stats analyses.R: R code containing statistical analyses used in this study.