Fisheries science often uses population models that assume no external recruitment, but nearshore marine populations harvested on small scales of <200 km often exhibit an unknown mix of self-recruitment and recruitment from external sources. Since empirical determination of self-recruitment vs. external recruitment is difficult, we used a modeling approach to examine the sensitivity of fishery management priorities to recruitment assumptions (self [closed], external [open]) in a local population of harvested giant clams Tridacna maxima on Mo’orea, French Polynesia. From 2006 to 2010, we measured growth, fecundity, recruitment, and survival (resulting from natural and fishing mortality). We used these data to parameterize both a closed (complete self-recruitment) and an open (no self-recruitment) integral projection model (IPM), and then calculated elasticities of demographic rates (growth, survival, recruitment) to future population abundance in 20 years. The models’ lowest projected abundance was 93.4% (95% CI, [86.5%, 101.8%]) of present abundance, if the local population is entirely open and the present level of fishing mortality persists. The population will exhibit self-sustaining dynamics (1 <= l <= 1.07) as for a closed population if the ratio of self-recruits per gram of dry gonad is >0.775 (equivalent to 52.85% self-recruitment under present conditions). Elasticity analysis of demographic parameters indicated that future abundance can most effectively be influenced by increasing survival of mid-sized clams (~80–120 mm) if the population is self-sustaining, and by increasing survival of juvenile clams (~40–70 mm) if the population is non-self-sustaining (as for an open population). Our results illustrate that management priorities can vary depending on the amount of self-recruitment in a local population.