Generally you'll be advised against it, however there are options on CPAN that attest to most things covered in your list. Heap::Simple is a slim and fast priority queue and other structures may fit other niches (ex. Tree::R, Graph). There's also of course PDL for vector programming. Lots of modules dealing with HTML and interchange formats (XML, etc.) implement trees internally. You have looked?

For general purpose though, no, basically use arrays for positional indexing and hashes for name-based indexing; map, grep and subs to operate on them (in place of classical ADT's) and classes (Moose) when you "need to go deeper."