Amphibian larvae exhibit phenotypic plasticity in size at metamorphosi
s and duration of the larval period. I used Pseudacris crucifer tadpol
es to test two models for predicting tadpole age and size at metamorph
osis under changing environmental conditions. The Wilbur-Collins model
states that metamorphosis is initiated as a function of a tadpole's s
ize and relative growth rate, and predicts that changes in growth rate
throughout the larval period affect age and size at metamorphosis. An
alternative model, the fixed-rate model, states that age at metamorph
osis is fixed early in larval life, and subsequent changes in growth r
ate will have no effect on the length of the larval period. My results
confirm that food supplies affect both age and size at metamorphosis,
but developmental rates became fixed at approximately Gosner (1960) s
tages 3 5-37. Neither model completely predicted these results. I sugg
est that the generally accepted Wilbur-Collins model is improved by in
corporating a point of fixed developmental timing. Growth trajectories
predicted from this modified model fit the results of this study bett
er than trajectories based on either of the original models. The resul
ts of this study suggest a constraint that limits the simultaneous opt
imization of age and size at metamorphosis.