A previous model of feeding by sea lamprey Petromyzon marinus predicted ene
rgy intake and growth by lampreys as a function of lamprey size, host size,
and duration of feeding attachments, but it was applicable only to lamprey
s feeding at 10 degrees C and it was tested against only a single small dat
a set of limited scope. We extended the model to other temperatures and tes
ted it against an extensive data set (more than 700 feeding bouts) accumula
ted during experiments with captive sea lampreys. Model predictions of inst
antaneous growth were highly correlated with observed growth, and a partiti
oning of mean squared error between model predictions and observed results
showed that 88.5% of the variance was due to random variation rather than t
o systematic errors. However, deviations between observed and predicted val
ues varied substantially, especially for short feeding bouts. Predicted and
observed growth trajectories of individual lampreys during multiple feedin
g bouts during the summer tended to correspond closely, but predicted growt
h was generally much higher than observed growth late in the year. This sug
gests the possibility that large overwintering lampreys reduce their feedin
g rates while attached to hosts. Seasonal or size-related shifts in the fat
e of consumed energy may provide an alternative explanation. The lamprey fe
eding model offers great flexibility in assessing growth of captive lamprey
s within various experimental protocols (e.g., different host species or th
ermal regimes) because it controls for individual differences in feeding hi
story.