Another one

In an order complete Banach lattice: is any finite-rank positive operator approximable by positive operators having their range in finite-dimensional sublattices?

Nielsen seems to say yes, but I could not repeat the proof.