DENSITY-DEPENDENT INTERACTIONS WITHIN A COMPLEX LIFE-CYCLE - THE ROLES OF COHORT STRUCTURE AND MODE OF RECRUITMENT

1. We analysed the effects of cohort structure, density, egg hatch inhibition and cannibalism on estimated per capita growth rate (r') in populations of Aedes triseriatus established in artificial habitats in Massachusetts. 2. As density increased from 0.5 K to 0.75 K to the estimated carrying capacity (K = 60 larvae per 100 ml), r' decreased, along with other measures of success: survivorship, female size, and development rate. 3. Cohort structure and recruitment schedule significantly influenced r'. Populations initiated as eggs achieved r' values greater than cohorts started as first instar larvae and substantially greater than those populations consisting of single cohorts. 4. We did not observe a significant relationship between hatch rate and larval density. In addition, hatch inhibition did not require direct larval contact with eggs. 5. We found no evidence for cannibalism in our experimental habitats, which were stocked with leaf detritus and treehole water, suggesting that refugia offered protection not gained in a simpler laboratory setting or that large larvae had sufficient levels of alternative food sources (Koenekoop & Livdahl 1986). 6. Egg-initiated cohorts exhibited the greatest developmental asynchrony during the first month of the experiment, whereas cohorts added as 2-day-old larvae displayed highly synchronized development. Increased competitive interactions in this latter treatment may explain the small size attained by adult females emerging from these populations. Differences among multiple cohort groups increased with increasing density. 7. Our results emphasize the need to incorporate overlapping stages into experimental designs for populations that develop asynchronously, and the value of using r' as a descriptor of success rather than the individual components of success (e.g. survivorship, size, and development rate) which can lead to misinterpretations of productivity.