The thesis presents research on the life-history polymorphism in the mite Sancassania berlesei. Males of this species are andropolymorphic: there are two distinct male phenotypes. One, the fighter, develops a third thickened leg pair, with which it kills off other fighters and males which do not exhibit a third thickened leg pair, the non-fighters. A review of the life-history of S. berlesei is given, focussing on its general biology, diet, dispersal and mating behaviour. This is followed by a review of the andropolymorphism, and the current understanding of the mechanisms underlying it. The major conclusions from the experimental work presented in this thesis are that fighters primarily develop at low population densities; though the proportion of males becoming fighters at any given density may change over time. This change is likely to be due to condition-dependence. Data is presented to illuminate these matters and a model is developed linking fighter development to the costs of being a fighter (in terms of survival) and the benefits of being a fighter (in terms of fecundity). The sex ratio in S. berlesei is 1:1, and there is no evidence of density or frequency-dependent deviations from this. A delay in food supply at maturation delays the time of maximum fecundity of females for about seven days and lowers their overall egg output. Density-dependent effects reduce the overall daily fecundity of females in higher densities. Female survival is affected by density, food present and rearing conditions. Nearly all eggs laid by S. berlesei hatch regardless of the conditions. Eggs laid in very poor conditions hatched even earlier than the average time of between day three and four. At density two, animals do synchronise their frequency, when isolated together from egg stage. Poor conditions reverse female density-dependence from convex to concave with the lowest life expectancy at intermediate densities. The trade-off between survival and fecundity is the likely cause. Amalgamating the results from the previous experiments, the influence of stochastic population dynamics on male strategy was then modelled. The results indicate that the fighter morph development rule is sensitive to the probability of low population densities arising. When low densities occur, there is a selective advantage to being a fighter. With increasing probability of lower densities, becoming a fighter is more feasible. The ESS rule changes, while in a stable high density environment a density-dependent fighter rule is never selected for. This indicates an influence of stochastic population dynamics on life-history evolution. Modelling demographic stochasticity in the fighter rule shows some buffering effect of this form of stochasticity. The fighter morph determination rule is less sensitive to environmental stochasticity with a high frequency of low densities. Using an agent based model with diploid genetics, I show that under high densities a fighter male is less successful at passing on his genes than a non-fighter. At a density of one male, the fighter gains no advantage to developing the fighter phenotype (as he is not competing with other males). In this case, the advantage may arise through future increases in density (such as through immigration or maturation of offspring). The density-dependent fighter development rule is then switched within the model from density-dependent to frequency-dependent, and the model indicates, that even under the frequency-dependent rule a possible ratio of fighters to non-fighters could exist. The system does not reach this state due to condition-dependence in reality. Following on from the findings discussed above, that morph determination has a condition-dependent component, I develop an argument that relates the observed forms of morph determination (density-dependent and frequency-dependent) in three closely related species of mites via an underlying condition-dependence. It is shown that condition-dependence is likely the linking factor between frequency and density-dependence. This is shown to be possibly a rule for all species displaying polymorphism which includes physical alterations of their bodies.