Feburary 8. Morphometrics

PAEDO- and PERAMORPHOCLINES

• arises if paedo- or peramorphic variant is better adapted to an adjacent environment on a cline
• many environments take the form of continuously variable gradient; e.g. shallow to deep water, coarse to fine grained sediment, high to low latitude, high to low temperature

EXAMPLE: PAEDOMORPHOCLINE Eocene brachiopod Tegulorhynchia

• finely ribbed (80 per shell)
• nearly spherical; with fold and sulcus
• pedicle foramen

All vary during ontogeny:

• ribs increase in number
• fold & sulcus deepens
• test becomes more spherical
• foramen gradually shrinks
• paeodomorphocline exists between Eocene and Miocene with successive species found in successively shallower, more energetic environments
• larger foramen in descendant species caused by neotenic slowing of plate growth
• decreased test convexity suggests smaller lophophore; larger lophophore is needed in deeper quieter water where organism must create own current to feed\
• PASSIVE or ACTIVE? several species coeval; little or no competition except at edges of ranges; passive spread into unoccupied niche space

EXAMPLE: PERAMORPHOCLINE tragelolaphids

MORPHOMETRICS

• each species represented in n-dimensional hypervolume whose coordinates are morphologic variates
• the axes' units can be standardized, log-transformed etc.
• principal components analysis (PCA) used to reduce dimensionality and discover orthogonality
• cluster analysis determines relative distance between species based on several morphological parameters

• assumption: arrangement of species in 'morphological space' accurately reflects their ecological relationships; this is not easily verifiable and is seldom tested

Example: desert lizards

• 10 morphometric variables: 1. snout-vent, 2. tail length, 3. body length, 4. width of head, 5. depth of head, 6. jaw, 7. forefoot, 8. forearm, 9. hindfoot and 10. hindleg length
• measurements strongly correlated with each other (r=0.75) among 92 spp.
• average morphometrics (for all samples of one species) used to represent species in 10 dimension morphospace
• each dimension given equal wt. by standardization
• hypervolume has "centroid" = hypothetical average lizard
• distances from centroid are calculated for each species and for averaged continental faunas

RESULTS:

• cluster analysis:

• Kalahari lizards more similar to each other than are North American or Australian lizards; overall average and 'nearest neighbor' distances smaller in Kalahari; hypervolume occupied by Kalahari lizards is 'compact'
• distances most variable in Australian lizards, perhaps caused by greater diversity
• principal components analysis:
• reduced original 10 variables to 3 components
• account for 96% of total variance:
• 1st component 77.7% "body size"
• 2nd component 11.6% "relative tail length"
• 3rd component 6.7% ??
• Australian thorny devil and American desert horned lizard converge in morphospace; both are ant-eating specialists