Variation in early post-settlement growth and mortality of scleractinian corals

For species with complex life histories, such as scleractinian corals, processes occurring early in life are likely to play a strong role in population regulation. A plethora of studies have examined settlement patterns of coral larvae, mostly on artificial substrata, and the composition of adult coral assemblages across multiple spatial and temporal scales. However, relatively few studies have examined the demography of small (≤ 50 mm maximum diameter) sexually immature corals on natural reef substrata, mostly due to difficulties associated with detecting small corals within natural reef environments. Mortality is often very high during the first year post-settlement, often reaching up to 99 %. However, if post-settlement mortality rates were universally high, then this would have limited influence on adult abundance. In fact, rates of juvenile mortality vary spatially, temporally and taxonomically, which will influence the structure and dynamics of populations and communities.\ud \ud In order, to better understand the influence of early post-settlement processes on patterns of abundance and community structure among adult corals, I first quantified the variation in abundance, composition and size of juvenile corals (≤ 50 mm diameter) among 27 sites, nine reefs and three latitudes spanning over 1000 km on Australia's Great Barrier Reef (GBR). A total of 2,801 juveniles were recorded with a mean density of 6.9 (± 0.3 SE) individuals per m², with Acropora, Pocillopora, and Porites accounting for 84.1 % of all juvenile corals surveyed. Size-class structure, orientation on the substratum and taxonomic composition of juvenile corals varied significantly among latitudinal sectors. The abundance of juvenile corals varied both within (6-13 individuals.m⁻²) and among reefs (2.8-11.1 individuals.m⁻²) but was fairly similar among latitudes (6.1-8.2 individuals.m⁻²), despite marked latitudinal variation in larval supply and settlement rates previously found at this scale. Furthermore, the density of juvenile corals was negatively correlated with the biomass of scraping and excavating parrotfishes across all sites, revealing a potentially important role of parrotfishes in determining distribution patterns of juvenile corals on the GBR. While numerous studies have advocated the importance of parrotfishes for clearing space on the substratum to facilitate coral settlement, my results suggest that at high biomass they may have a detrimental effect on juvenile coral assemblages.\ud \ud I next examined the effects of incidental grazing on post-settlement survival of Acropora cytherea recruits. Larvae of A. cytherea were reared in captivity and settled onto terracotta tiles. Replicate tiles were deployed within the exposed reef crest and the sheltered back reef of Lizard Island, in northern section of the GBR. Overall, survivorship was broadly comparable between habitats, ranging from 37.7 – 64.5 % per month on the exposed reef crest, and 53.1 – 64.3 % on the sheltered back reef. On the reef crest, the exclusion of herbivores increased survivorship by 22.4 %: from 42.1 to 64.5 % per month. Moreover, recruits survivorship within the reef crest was negatively correlated with the density of parrotfish feeding scars on tiles after 4-weeks. In contrast, the exclusion of herbivores had no detectable effect on survivorship within the back reef and no feeding scars were observed on tiles in this habitat. Difference in grazing-induced mortality between habitats was most likely related to differences in herbivore size and abundance, with parrotfish biomass being 5.5-fold greater on the reef crest than the back reef. Surprisingly, tile orientation had no effect on survivorship of A. cytherea in either habitat, despite a marked difference in the sediment cover on vertical (0 %) vs. horizontal tiles (30 %) in the back reef. This was in marked contrast to previous studies that have reported sedimentation is a major cause of early post-settlement mortality in corals.\ud \ud Ultimately, I wanted to directly measure the fate of juvenile corals on natural substratum, from small (within transect) to large scales (among latitudinal sectors of the GBR, and geographic comparisons between the GBR and French Polynesia), to better understand the role of post-settlement processes in shaping adult communities. Here, I measured growth and mortality rates of Acropora, Pocillopora and Porites juveniles within quadrats, over a period of 18 months. On the GBR, these processes were significantly different among the three taxa examined, and the highest variation occurred among quadrats within transect. In particular, mortality of Acropora juveniles (mean of 29.3 % per annum) was lower than for Pocillopora (33.6 % per annum) and Porites (43.2 % per annum); and the growth rates of Acropora (mean of 1.80 mm.month⁻¹) and Pocillopora (mean of 1.76 mm.month⁻¹) were higher than for Porites (mean of 0.97 mm.month⁻¹). Latitudinal patterns were not consistent among taxa; highest mortality was recorded within the central sector of the GBR for Acropora juveniles and lower mortality at the northern sector for Porites juveniles, whereas for growth rates, only those of Pocillopora juveniles varied at sector scale, with lower growth in the southern sector. Mortality rates decreased as the size of juveniles increased. In addition, mortality rates were positively correlated with adult cover for Acropora and Pocillopora juveniles probably due to competition for space, with adult overgrowing juveniles. However, mortality rates were not correlated with parrotfish biomass, probably due to the ability of the fish to avoid juveniles, as opposed to very small recruits. In conclusion, generic-level assemblage structure of corals on the GBR is driven largely by high variations in juvenile mortality and growth rates at small and large-scales.\ud \ud When comparing post-settlement rates from Trunk reef in the central GBR with those in Moorea, French Polynesia, I found that there were no differences in the density, growth, or mortality rates of juvenile corals. However, significant regional differences exist in the taxonomic composition of coral assemblages within both adult and juvenile assemblages, with Pocillopora being the predominant coral genera in Moorea and Acropora at Trunk Reef. Most of the variations in these variables were evident at the small (within reef) scale, with exposed sites having lower densities and higher rates of mortality of juvenile corals than sheltered sites at both locations. The lack of geographic variation in the density, growth and mortality rates of juvenile corals is interesting given the cover of adult coral was 3-fold higher on Moorea (31.1 %) than on Trunk Reef (10.8 %), suggesting that adult coral assemblages are structured more by differential adult mortality, larval settlement, or very early post-settlement mortality (before colonies can be observed in situ), rather than demographic rates of juvenile growth or mortality