In this paper I revisit the connection between edge percolation and the col
lapse transition in lattice animals. It was shown by Domb (1976 J. Phys. A:
Math. Gen. 9 L141) that the critical percolation point is a theta-transiti
on in a certain model of self-interacting animals. I extend this result by
showing that the free energy of lattice animals in the cycle-contact model
is non-analytic in contact- and cycle-activities at points other than the c
ritical percolation point. This correspondence between percolation and coll
apsing animals suggests that the collapse transition in this model may be r
elated to percolation. Contact-collapse in animals is then studied from a p
ercolation of clusters of contacts perspective. I first argue that there is
a critical activity for percolation of clusters of contacts, and then inve
stigate this numerically.