Supplementary Information to van Hinsbergen et al., PLoS One (submitted, 2026)

Supplementary Files 1: GPlates files (www.gplates.org, [14]) of the Utrecht Paleogeography Model version 1.0 (2025) presented in the paleogeographic maps of Figures 1 and 3, and the rigid polygon version that is used as basis for the Paleolatitude.org tool. These consist of a rotation file, and a series of shape files (in gpml format) that underpin the paleogeographic model, a gpml file of the rigid polygons that are used to rotate coordinates in the Paleolatitude.org tool, as well as a project file (.gproj) of the entire paleogeographic reconstruction. Supplementary Files 2: Details of gAPWP25. In addition to a Readme.txt file with general descriptions and the gAPWP25.rot file with the updated paleomagnetic reference frame in GPlates rotation format, the files contain: Table S1: Changelog of the update of gAPWP23 to gAPWP25 Table S2: Paleomagnetic database used to compute the global apparent polar wander path for the last 320 Ma. We have listed age constraints, statistical parameters, Euler rotation parameters and other metadata per paleomagnetic pole used in the parametric re-sampling scheme. For more details, see main text. The grey-colored entries are excluded from the computation of the APWP. See columns 'age constraints', 'comments' and 'reliability' for specific details for a given dataset. Abbreviations: min_age and max_age = lower and upper boundaries of age uncertainty range; slat/slon = latitude and longitude of (mean) sampling location; N = number of paleomagnetic sites used to compute the paleopole; mDec/mInc = mean declination of inclination; α95/A95 = radius of the 95% confidence circle about the mean of the distribution of directions/VGPs; k/K = Fisher [107] precision parameter of the distribution of directions/VGPs; plat/plon = paleopole latitude and longitude (south pole); K_est/A95_est = values estimated using formula of Cox [211] (eq. 24); plateID = plate identification number; Rlat/Rlon = paleopole latitude and longitude in coordinate frame of South Africa; EP_lat/EP_lon/EP_ang = total reconstruction pole parameters for rotating the paleopole to South Africa coordinates; f = flattening factor (only for sedimentary data), p_std = standard deviation of the assumed normal distributed co-latitudes, obtained from E/I correction (only for sedimentary data); Deenen = indicates whether the N-dependent reliability envelope of Deenen et al. [212] is satisfied (TRUE or FALSE) or, in case of sediment-derived datasets, the quality grade (A, B or C) following the evaluation scheme of Vaes et al. [104]; excl = reason for exclusion (R = rejected because entry is a duplicate, N < 5, age range > 20 Ma, remagnetized or otherwise considered unreliable, see comments/reliability column); refno = reference number in global paleomagnetic database [213, 214]; DB = database in which entry is listed (T12 [27], PSV10 [215], gAPWP23 [13], gAPWP25 = added in this study). Table S3: Global plate circuit used to transfer paleomagnetic data to a single reference plate, from Vaes et al. [13] with minor modification in the rotation parameters for India. See Vaes et al [13] for references and details. Table S4: gAPWP25 rotated in the coordinates of the major continents. Supplementary Files 3 Table S5: Euler rotations of every polygon relative to South Africa (701) at times corresponding to the ages of the reference poles of gAPWP25, using the rotation file of the Utrecht Paleogeography Model in Supplementary Files 1. Paleolatitudes provided by Paleolatitude.org 3.0 are interpolated from these rotations.