Well, not quite. The first problem here is that "je" combines the klama (which you typoed, btw) and sisku into a single entity, x1 is a go-and-seeker of/goes-and-seeks x2. While x2 for compound tanru like that are not actually defined, it often is taken to mean that the x2 should be able to fit into the x2 of both parts (the klama and the sisku). Which implies that he is a goer-to and seeker-of ballness... Which is a bit messy. "gi'e" is cleaner here:
ma'a klama gi'e sisku tu'a lo ko'a gercmo . This puts the "tu'a lo ko'a gercmo" associated only with the sisku, not the klama. But in fact, I would drop the "klama gi'e" entirely. Because the English "go and" is kind of just a grammatical fillip. It doesn't really mean to go someplace in particular. It just emphasizes a course of action.
Now, you will also notice I used "tu'a lo ko'a gercmo" instead of "lo ka ko'a gercmo". There is two reasons I did this. First, I do prefer to use tu'a X when seraching for an object (as it says in jbovlaste). Although this may have changed in modern day usage. I'm a dinosaur. And secondly, because "lo ka ko'a gercmo" does not mean " the property of his bark", but in fact means "the property of him being a barker" (because the ko'a is now part of the bridi "ko'a gercmo" (he barks), not a possessive like it was before, without the ka.
--gejyspa