Am I right in thinking that the only way to test each individual sub would be to create a module eventhough if the code is not re-usable by any other scripts (as suggested by Thilosophy)?

All you have to do to test individual subs is to load a pile of code somehow and control the flow of execution from something other than the program itself.

Granted, this is easier if you create one or more modules, but it's not a requirement. (The purpose of modules is to help you organize code, not just to reuse it.)

Am I right in thinking that the only way to test each individual sub would be to create a module eventhough if the code is not re-usable by any other scripts

No. There are as many other ways to test as your imagination can dream up.

You could for example, use a command line switch that when supplied causes a different main-like sub to be invoked in place of main.

...
if( $DEBUG ) {
    debug_main()
}
else{
    main();
}

## The rest of the code
[download]

And run the tests using perl -s TheScript.pl -DEBUG

Or using Smart::Comments something like:

...
### debug_main()
### exit;
main()
# the rest of the code
...
[download]

Now you run your tests using perl -MSmart::Comments TheScript.pl

And within debug_main() you can use any testing tools that you want. You could, for example, use the traditional ok(..) and not_ok( ... ), and then use the test harnesss or prove commands with an appropriate command line switch to perform white box testing.

You could even combine the two. Rather than having to structure your application to suit the test tools, you can structure your application in whatever way best suites its requirements, and use whichever combination of test methodologies best suit your needs.