1 Supplementary Information: Materials for Radiometric Dating 2 3 Consistent with all other studies of island evolution [e.g. Woodroffe et al., 2007; Kench et al., 4 2012; Kayanne et al., 2012] we radiometrically dated carbonate material to provide a 5 temporal framework for the deposition of island sediments on Jabat. Radiometric dating of 6 skeletal carbonate sediments can be problematic as a disparity may exist between the time of 7 death of the organism and time of final deposition, which may occur after an undefined 8 period of transport, breakdown, and mixing. In such instances, samples yield the earliest 9 possible time of deposition. Consequently, interpretation of island accumulation using this 10 dating approach must be undertaken with caution. Careful selection of samples, with respect 11 to the degree of abrasion, is necessary to minimise the time lag between organism death and 12 deposition and in order to evaluate which sediments may be more prone to time lags. 13 In this study three types of skeletal materials were selected for analysis. Each type of skeletal 14 material has different reliability with respect to the temporal lag between organism death and 15 deposition. 16 17 First, a set of five in situ coral samples were selected for analysis. These samples were 18 retrieved via drilling of the reef flat surface and base of cores (Figure 2b, S3). The reef flat 19 samples were interpreted as in situ corals based on the ‘intact’ condition of corallites, the 20 orientation of corallite structures, orientation of the growth form within cores, and clear and 21 horizontal boundary between the coral and underlying basement (Fig. S2). Dates on these 22 corals provide an accurate indicator of the death of the organism and the in situ condition 23 ensures there has been no post-mortem transport. In this study ages on in situ corals provide a 24 temporal framework on final stages of reef growth. 25 26 Second, seven cobble-size coral clasts were dated. In each case cobbles had intact corallite 27 structures and clasts were not rounded. As shown by Ford and Kench [2012] corallite 28 structures are readily worn down in the initial stages of transport in littoral systems. Such 29 characteristics indicate the clasts have not undergone significant post-mortem littoral 30 transport. Furthermore clasts dated showed an absence of secondary erosion and boring as 31 may be expected if cobbles had been lying inert on the deeper fore reef for a considerable 32 period of time between death and transport onto the reef platform. Consequently, it is 33 thought that deposition of these clasts occurred rapidly in a high energy event or sequence of 34 events and that the time between organism death and deposition is short. In this study ages on 35 the coral cobbles provides an age framework for deposition of the underlying gravel 36 basement within the island. 37 38 Third, 13 ages on sand-size materials were obtained. Woodroffe et al. [2007] review the 39 difficulties in dating sand-size skeletal grains and concluded that single constituent dating 40 may provide a more reliable indicator of the time of deposition. In their study Woodroffe et 41 al. [2007] used foraminifera and molluscan grains to provide a chronological account of the 42 accumulation of Warraber Island. The premise of this approach is to use those components 43 within deposits that appear to be intact and have undergone very little abrasion. However, the 44 selection of single grains can also lead to considerable error in defining a time of deposition. 45 Samples selected for analysis in this study were rich in the foraminifer Calcarina sp. These 46 components were preferentially selected for dating as they preserved near intact tests as 47 opposed to coral and coralline algal grains. Of the three types of material chosen for dating it 48 is expected that temporal lags would be more apparent for the sand-size sediments. It is 49 important to stress that ages on sand-size materials are used in this study to define a general 50 window within which the accumulation of sand-size materials occurred on Jabat, rather than 51 provide a definitive chronology of sedimentation. However, several pieces of evidence 52 suggest that the ages can be used to construct a maximum age of accumulation of sand-size 53 sediments within the island. First, foraminifera selected for analysis generally had near 54 pristine tests. Second, results reveal a consistent ages and a lack of inversions. 55 56 References 57 Ford, M. R. and P. S. Kench (2012), The durability of bioclastic sediments and implications 58 59 for coral reef deposit formation, Sedimentology, 59(3), 830-842. Kayanne, H.T., T. Yasukochi, T. Yamaguchi, and H. Yamano (2011), Rapid settlement of 60 Majuro atoll, central Pacific, following its emergence at 2000 years Cal BP, Geophys Res 61 Let, 38, L20405, doi:10.1029/2011GL049163. 62 Kench, P.S., S.G. Smithers, and R.F. McLean (2012), Rapid reef island formation and 63 stability over an emerging reef flat: Bewick Cay, northern Great Barrier Reef, Australia, 64 Geology, 40, 347–350. 65 Woodroffe, C.D., B. Samasorn, Q. Hua, and D. Hart (2007), Incremental accretion of a sandy 66 reef island over the past 3,000 years indicated by component-specific radiocarbon dating, 67 Geophys Res Lett, 34, L03602, doi:10.1029/2006GL028875.