In situ each sample was inspected and those fragments without evidence of bleaching, dead tissue or a high amount of sponge coverage were selected and re-attached within the same area of collection using the plastic cable tie method (Edwards, 2010) on both artificial and natural substrates. Stable coral rubble was used as natural substrata (Fig. 2c) for all four study sites with depths of 3e8 m: Zona de Restauración (n ¼ 20), Cueva del Muerto (n ¼ 19), Amarradero (n ¼ 12), Plataforma Pavonas (n ¼ 21). Only at Zona de Restauración artificial underwater structures were installed because fragments were available in sufficient numbers for restoration. Artificial concrete structures (modules of 60 cm2) (n ¼ 10) had w12 steel-stacks each and were stabilized at 3e4 m depth (Fig. 2a). In both substrates each fragment was identified, measured and tagged. Coral growth progress was measured monthly (cm mo1) during one year; using calipers (0.05 mm precision) noted maximum apical height growth distance measured from bottom to top of thefragment and maximum radial growth, referred as the longest diameter perpendicular to the fragment height. Coral fragments
were firmly adhered on the substrate were recorded as Attachment Rates; live fragments with a continuous growth over the entire experiment were recorded as Survivorship Rates and those fragments which died in place during the study were recorded as Mortality Rates.