Population consequences of mutual attraction between settling and adult barnacles

1. Spatial patterns of recruitment were compared between populations of the rocky shore barnacles (Crustacea: Cirripedia) Semibalanus balanoides, an obligate cross-fertilizer, and Chthamalus montagui, which can self-fertilize. We tested the hypothesis that recruitment depends on a behaviourally mediated interaction at settlement between the effects of adult background density and adult spacing, which limit free space for settlement and mating opportunities, respectively. Recruitment to patches of cleared rock (10- or 30-cm diameter) was compared between replicate shores with background densities of adult barnacles classed as low (less than or equal to20 per 25 cm(2)) and high (>4x low). Replicate patches were cleared of all barnacles surrounding a remnant cluster, comprising 0, 1, 2, 4, 8, 16 or 32 adults. 2. For S. balanoides, settlement and subsequent recruitment over 5 months varied in direct proportion to remnant cluster size, except on the shore with the highest background density where recruitment was inversely proportional to cluster size. We interpret this inversion to indicate attraction to potential mates at low adult densities (positive density dependence, an Allee effect) switching to attraction to free space at high density (negative density dependence). The strengths and slopes of the regressions increased from shores with the lowest to the highest overall recruitment of barnacles, but retained significance over a five-fold range in recruitment. Positive effects of cluster size on recruitment were consistent between consecutive years, despite considerable variation in recruitment densities. In contrast, recruitment of C. montagui was generally more weakly proportional to cluster size, except for a strong positive correlation at the shore with the highest recruitment. 3. Dispersion of recruits within treatment patches was accurately modelled by a computer simulation that allowed each barnacle to settle at random between fixed minimum and maximum distances from the nearest other settled barnacle. The model estimated threshold distances by maximum likelihood fit to observed recruitment into concentric annuli around the central adult cluster. Upper thresholds of separation corresponded to penis length for 65% of S. balanoides and 42% of C. montagui patches. Lower thresholds were similar to2x cyprid length for 75% of S. balanoides patches, but were larger than this for C. montagui patches.