Why morphometrics? Measuring biological variation

Transcription

1 Geometric morphometrics as a useful tool for visualising and analysing deformities in fish D. Adriaens1, Y. Verhaegen1, T. De Wolf2, P. Dhert3 & P. Sorgeloos Evolutionary Morphology of Vertebrates (UGent, Belgium) Maricoltura Rosignano Solvay (Italy) INVE Technologies (Dendermonde, Belgium) Artemia Reference Centre (UGent, Belgium) dominique.adriaens@ugent.be

2 Workshop "Deformities in Fish Larvae" 2/22 Why morphometrics? Measuring biological variation growth - development interspecific variation species-specific specific features intraspecific variation phenotypic plasticity deformations

3 Traditional morphometrics Length measurements advantages easy to measure easy to analyse PCA, DFA,... a b disadvantages size variables, not shape variables unsufficient model of shape a b March a=a b=b shapes are equal Workshop "Deformities in Fish Larvae" 3/22

4 Traditional morphometrics Length ratios advantages standardised, size removal a b e disadvantages reduces variation size removal incomplete always dependent of size size removal erroneous c d a b e a a = standard length b=b b/a b /a c=c c/a c /a d=d d/a d /a e=e e/a e /a March d shapes are completely different c Workshop "Deformities in Fish Larvae" 4/22

5 Solution to the problems? Geometric Morphometrics

6 Workshop "Deformities in Fish Larvae" 6/22 Data types Outlines advantages very good model of true shape especially rounded shapes mathematical model shape parameters statistics allowed PCA, DFA,... disadvantages not good for certain shapes shape changes within the outline model shapes with pointed outlines no link to actual shape variation in specimens

7 Workshop "Deformities in Fish Larvae" 7/22 Data types Landmarks advantages good model of true shape homologous points also shapes within shapes mathematical model shape parameters statistics allowed PCA, DFA,... deformation grids visualisation of shape differences disadvantages not good for certain shapes rounded shapes

8 Landmark-based morphometrics Thin Plate Splines data Cartesian coordinates 2D, 3D Y Y X March Workshop "Deformities in Fish Larvae" 8/22 X

9 Landmark-based morphometrics Thin Plate Splines data Cartesian coordinates 2D, 3D standardisation Generalised Procrustes Analysis size rotation position only true shape remains Y Y X March Workshop "Deformities in Fish Larvae" 9/22 X

10 Workshop "Deformities in Fish Larvae" 10/22 Thin Plate Splines Shape variation decomposition reference shape consensus new shape variables partial warps uniform shape variation compression shear non-uniform shape variation overall deformation localised deformation partial warp scores weight matrix

11 Workshop "Deformities in Fish Larvae" 11/22 Thin Plate Splines Shape variation analysis Relative warp analysis PCA on partial warp scores visualisation of shape variation min. RW1 max. RW1 min. RW1 RW2 RW1 max. RW1 DFA on partial warp scores between-group shape differences

12 Geometric morphometrics Application for studying deformations in Sparus aurata

13 Workshop "Deformities in Fish Larvae" 13/22 Data acquisition Digital images specimens n=40 Landmark digitisation number of landmarks 26 (type 1 & 2) GPA consensus specimens

14 Workshop "Deformities in Fish Larvae" 14/22 Data analysis Partial warp decomposition uniform variation no clustering Deoperculated Normal non-uniform variation partial warp 9 normal and deoperculated cluster normal less variation partial warp 12 normal and deoperculated cluster normal less variation

15 Workshop "Deformities in Fish Larvae" 15/22 Data analysis Normal Relative warp analysis RW1 (60.25%) Deoperculated

16 Workshop "Deformities in Fish Larvae" 16/22 Data analysis Normal Relative warp analysis RW2 (9.37%) Deoperculated

17 Workshop "Deformities in Fish Larvae" 17/22 Data analysis Relative warp analysis RW3 (7.38%) Normal RW4 (4.67%) Deoperculated

18 Workshop "Deformities in Fish Larvae" 18/22 Trends in shape changes Normal to deoperculated shape changes Normal Deoperculated Normal Deoperc. not size related distrupt variation developmental treshold? early onset of deformation?

19 Workshop "Deformities in Fish Larvae" 19/22 Trends in shape changes Normal shape changes min Normal Deoperculated max. not size related gradual variation phenotypic plasticity? deformities?

20 Trends in shape changes Deoperculated shape changes min Normal Deoperculated max. not size related gradual variation March phenotypic plasticity? variation larger than normal different than normal phenotypic plasticity Workshop "Deformities in Fish Larvae" 20/22

21 Workshop "Deformities in Fish Larvae" 21/22 Left-right asymmetry Specimen C1-8 right - #30 normal left - #9 strongly deoperculated right left

22 Workshop "Deformities in Fish Larvae" 22/22 Conclusions Use of geometric morphometrics more solid shape descriptors more correct standardisation (GPA) statistical analyses applicable visualisation of shape variation Deformations in Sparus aurata geometric morphometrics very useful shape-gap between normal and deoperculated within-group variation not size related extreme left-right asymmetry