Trematodes in their first intermediate mollusc host are limited both in space and in resources, which can place these parasitic flatworms under selective pressure to defend their habitat from competitors. In extreme cases this can lead to division of labor, with some worms specializing to defend against other parasites. These “soldiers” bolster the fitness of their clonemates by preventing resource loss or death from competitors attempting to invade their host. As a first step toward understanding how trematodes form the soldier caste, I developed a technique for estimating the death rates of individual worms within the clone. Chapter 1 shows how an understanding of the developmental biology of trematodes permitted me to estimate the lifespan of senescing worms. Coupled with a careful census of the clone, this permitted a calculation of within-host death rates. In Chapter 2, I developed a stage-structured projection matrix model of within-host dynamics to quantify how trematodes in the family Philophthalmidae produce the soldier group. The solved model estimated every rate governing within-host dynamics. I showed that philophthalmid soldiers (and likely those of all other previously known “social” trematodes) are developmentally arrested, ontogenetically juvenile worms capable of transitioning to the reproductive stage. The model also revealed how trematodes that produce soldiers via developmental arrest differ from a hypothetical species that possessed a truly non-reproductive, sterile caste. Chapter 3 documents the discovery of a trematode with such a sterile caste. Haplorchis pumilio is a globally invasive, human-infectious trematode that has established in southern California. Its soldiers are likely truly permanent, as they seem to follow an irreversible developmental trajectory that does not lead to the reproductive stage. This caste comprises the most morphologically specialized trematode soldiers yet described, with a pharynx volume 5 times greater than that of reproductive clonemates. Its soldier caste has elevated H. pumilio to the dominant competitive position in its North American invasive guild. In areas where trematodes are highly interactive, I estimated that 90% or more of potential trematode diversity was lost due to H. pumilio killing competitors.