Main content area

A comparison of epibenthic reef communities settling on commonly used experimental substrates: PVC versus ceramic tiles

Mallela, J., Milne, B.C., Martinez-Escobar, D.
Journal of experimental marine biology and ecology 2017 v.486 pp. 290-295
Scleractinia, algae, benthic organisms, ceramics, community structure, coral reefs, corals, data collection, ecosystems, lawns and turf, models, multivariate analysis, pipes, tiles, Great Barrier Reef
Artificial substrates are routinely used in coral reef research to model the recruitment and growth responses of benthic organisms (e.g. coral recruitment and encrusting organisms) to environmental change. Two commonly used, but structurally different, artificial substrates include cylindrical PVC pipes and flat ceramic tiles. Various ecosystem based models extrapolate data from these substrates interchangeably based on the assumption that results are directly comparable. In order to test this assumption we deployed these commonly used artificial substrate materials, PVC poles and ceramic tiles, in shallow patch reefs for 34months at One Tree Island, Great Barrier Reef. Tiles were positioned to mimic upwards facing, well-lit substrates (exposed), and downwards facing, shaded (cryptic) substrates. Multivariate analyses demonstrated that the community composition differed significantly between all three treatments. The majority of artificial substrate, coral reef experiments focus on key groups of calcifying organisms, primarily: coralline algae, scleractinian coral and/or total calcareous encruster cover. Interestingly, significant differences in the recruitment, colonisation and community composition of these organisms were detected for our three treatments. When compared to ceramic tiles, PVC poles had greater coverage of crustose coralline algae but reduced levels of coral recruits (<1mm diameter) and turf algae. We suggest that comparisons between studies that utilise data from different substrate types should be used with caution. Additionally, large scale modelling and forecasting exercises utilising these data sets should adjust for the inherent biases of each method.