| Abstract | In the recent decades, marine benefits lost due to marine biodiversity and coral reef communities are globally and seriously threatened by anthropogenic impacts. Many people and programs from different levels have involved in coral reef management in various method approaches. However, practical coral reef management has been facing with several difficult matters, one of these matters is lack of adequacy biological monitoring tools and indicators. Using the causality framework based on the relationship between pressure of anthropogenic activities and ecological coral reef state and their association components functions, the main goal of this study is to constitute a Coral Reef Index of Biological Integrity (CR-IBI) to support a part of coral reef management strategy. Based on the principles of IBI development process that is requirement the set of "Reference" and "Degraded" sites separately. Unavailable environmental data and criteria regard to chemical and physical measurement parameters, the criteria to determine coral reef "Reference" and "Degraded" sites are based on site substrate components and structures recorded from 3023 transects/sites by ReefCheck monitoring program from 1997 to 2004. With criteria of "Reference" site are: more than 35% of hard coral frequency of occurrence, the ratio of {(Fleshy algae +Dead coral)/(Fleshy algae + Dead coral + Hard coral)} <35%, and Hard coral/(Fleshy algae + Dead coral) >_ 2, conversely, criteria of "Degraded" site are: percentage of hard coral frequency occurrences <30%, {(Fleshy algae +Dead coral)/(Fleshy algae + Dead coral + Hard coral)} >50%, and Hard coral/(Fleshy algae + Dead coral) _< 1, 76 reference sites and 72 degraded sites classified in shallow environment, while 39 reference sites and 37 degraded sites classified in deep environment. Ecologically and naturally, from ten global ReefCheck bioindicators fifty-five candidate metrics of the index compiled and defined based on Fishbase, ReefCheck database and other sources were identified. Of which, eight metrics in shallow environment and four metrics in deep environment, that can functionally reflect significant difference between reference and degraded condition, are selected to combine into final indices. Because of quite different distribution pattern of metric values among metrics, so criteria for metric value thresholds are different from metric to metric. The index is calculated by scoring each selected metric as 5, 3 and I depending on metric value threshold criteria. In calibration data set, the overall site classification efficiencies are low 65.97% in the shallow environment and 66.13% in deep environment. Applicability of index in narrative ratting site condition is not absolutely corrected site-to-site. However, the strongly negative correlation between index with "dynamite fishing" (-0.286**) and "number of yacht within 1 km" (-0.185*) in shallow environment and with "poison fishing" (-0.279*) and "coral damaged by other factors" (-0.283*) in deep environment indicate that coral reef indices are significant responses to stressors and promise reliable applicability. Key tivords: Index of biological integrity, Bioindicator, "Reference" and "Degraded " sites, Attribute, Metric, Threshold, Score, Stressor and Classification Efficiency IIIAbstract In the recent decades, marine benefits lost due to marine biodiversity and coral reef communities are globally and seriously threatened by anthropogenic impacts. Many people and programs from different levels have involved in coral reef management in various method approaches. However, practical coral reef management has been facing with several difficult matters, one of these matters is lack of adequacy biological monitoring tools and indicators. Using the causality framework based on the relationship between pressure of anthropogenic activities and ecological coral reef state and their association components functions, the main goal of this study is to constitute a Coral Reef Index of Biological Integrity (CR-IBI) to support a part of coral reef management strategy. Based on the principles of IBI development process that is requirement the set of "Reference" and "Degraded" sites separately. Unavailable environmental data and criteria regard to chemical and physical measurement parameters, the criteria to determine coral reef "Reference" and "Degraded" sites are based on site substrate components and structures recorded from 3023 transects/sites by ReefCheck monitoring program from 1997 to 2004. With criteria of "Reference" site are: more than 35% of hard coral frequency of occurrence, the ratio of {(Fleshy algae +Dead coral)/(Fleshy algae + Dead coral + Hard coral)} <35%, and Hard coral/(Fleshy algae + Dead coral) >_ 2, conversely, criteria of "Degraded" site are: percentage of hard coral frequency occurrences <30%, {(Fleshy algae +Dead coral)/(Fleshy algae + Dead coral + Hard coral)} >50%, and Hard coral/(Fleshy algae + Dead coral) _< 1, 76 reference sites and 72 degraded sites classified in shallow environment, while 39 reference sites and 37 degraded sites classified in deep environment. Ecologically and naturally, from ten global ReefCheck bioindicators fifty-five candidate metrics of the index compiled and defined based on Fishbase, ReefCheck database and other sources were identified. Of which, eight metrics in shallow environment and four metrics in deep environment, that can functionally reflect significant difference between reference and degraded condition, are selected to combine into final indices. Because of quite different distribution pattern of metric values among metrics, so criteria for metric value thresholds are different from metric to metric. The index is calculated by scoring each selected metric as 5, 3 and I depending on metric value threshold criteria. In calibration data set, the overall site classification efficiencies are low 65.97% in the shallow environment and 66.13% in deep environment. Applicability of index in narrative ratting site condition is not absolutely corrected site-to-site. However, the strongly negative correlation between index with "dynamite fishing" (-0.286**) and "number of yacht within 1 km" (-0.185*) in shallow environment and with "poison fishing" (-0.279*) and "coral damaged by other factors" (-0.283*) in deep environment indicate that coral reef indices are significant responses to stressors and promise reliable applicability |