Connell asked in 1985 whether initial settlement or post-settlement mortality determines the abundance of marine populations. We have investigated this question for abalone. Post-settlement effects will be most important if they are strongly density-dependent (DD), so that they damp out settlement variation. Studies of DD mechanisms over all life stages are very scarce. We estimated responses in greenlip abalone (Haliotis laevigata), from settlement to maximum size, using experiments on abalone farms and in the field. There is a short larval stage, a post-larval stage on exposed algae, a cryptic juvenile stage under boulders, and then they emerge into seagrass areas at maturity. We altered larval densities in replicate tanks to ask if this determined settlement rate. We manipulated densities of recently settled abalone on algal covered plates, as well as food supply, and measured growth and survival. We built replicate reefs of natural boulders at 8 field sites and seeded them with different densities of juveniles, then measured growth, and estimated mortaltiy using a tag-recapture model. We tagged adults, then reduced their densities in three areas paired with control areas, to examine effects on growth and fecundity.

Results/Conclusions The % of larvae settling appeared to be independent of density. In contrast post-larval survival and growth both decreased strongly with density, as did the growth of juveniles, and slower growth led to later maturation.. Survival of juveniles (after initial mortality, from 3-9 months afer seeding) was weakly DD. Natural mortality of adults is known to be low. Adults at the asymptotic size did not grow further at lower density, but increased gonad sizes. Growth increased progressively in smaller adults, and females showed increased gonad size in inverse proportion to somatic growth, while all males increased testis size. The strongest overall effect is DD growth, which we suggest is due to competition for food. This effect will shorten the effective generation time when density is reduced, plus reduce total mortality between birth and maturity. While the effects of DD mortality and fecundity have been well explored, the importance of DD growth has only recently received attention. Our simulations of populations show strong compensatory DD growth alone can stabilise populations and maintain abundance. Further, in harvested populations DD effects are necessary for sustainability under consistent fishing. If strong DD growth occurs our simulations show that standard stock assessment models would be biased so as to increase the risk of fishery collapses.