Reply to comment by J. G. Meert and R. Van der Voo on ‘New palaeomagnetic result from Vendian red sediments in Cisbaikalia and the problem of the relationship of Siberia and Laurentia in the Vendian’

Abstract

We thank J. Meert and R. Van der Voo for their attention to our paper. We are also glad that it has reopened a discussion about latitudinal position of Laurentia in Late Vendian, which was one of the aims of our publication. Unfortunately, Meert and Van der Voo partly misinterpreted our conclusions. Our aims were to present a working hypothesis that explained the most reliable palaeomagnetic data. The principle of multiple working hypotheses permits us to present a new model without compelling us to restate all of the arguments used to support previous models. We have decided not to accept some palaeomagnetic results that are central to the existing model of a polar Vendian position of Laurentia (Meert et al. 1994; Torsvik et al. 1996). On the other hand, this widely accepted model dismisses important data that support our new interpretation. In particular, many of our differences are based on the acceptance or rejection of palaeomagnetic baked-contact and fold tests. The Sept-Iles ‘A’ pole (Tanczyk et al. 1987) is the only one among coeval Laurentian poles that has a full contact test, unlike Catoctin ‘A’ (Meert et al. 1994) or the Callander pole (Symons & Chaisson 1991). It is very hard to explain in this case why the magnetization of the main intrusion is younger than the magnetization of the cutting dykes (pole ‘B’), which was suggested by Meert and Van der Voo in this forum and earlier publications. The result of the contact test in the Callander Complex described by Symons & Chaisson (1991) can be also explained by a regional bipolar remagnetization; the remanence direction of the country rocks is almost antiparallel to that of the studied dyke and close to the negatively inclined remanence direction of some other sampled sites of the Complex (sites 5, 6, 10 in Table 1 of Symons & Chaisson 1991). Of course, a primary magnetization is possible too. So, our conclusion was that at present it is very hard to propose a model that will fit both Callander and Sept-Iles ‘A’ results, and that the palaeomagnetic community should equally consider accepting either one of those results. Meert and Van der Voo also insist that their Catoctin ‘A’ pole is primary, and Catoctin ‘B’ is secondary (Meert et al. 1994). We think that Meert’s and Van der Voo’s claim that the fold test for the A component is positive at the 92 per cent confidence level is less than rigorous. Although somewhat arbitrary, the 95 per cent confidence level has been universally accepted as a fair benchmark of statistical significance. Although we agree with Meert and Van der Voo that unfolding leads to better grouping, in all fairness they must report their fold test as statistically inconclusive. Their discussion has, nonetheless, prompted us to re-examine their original data more carefully (Table 2 and Fig. 7 in Meert et al. 1994), and we note that the increase of Catoctin ‘A’ grouping after the tilt correction is due to just one site (site 3). Without this site the grouping before and after the tilt correction are statistically the same. Taking the data at face value, then there are only two high-quality poles for Laurentia between 580 Ma and 560 Ma: 565 Ma Sept-Iles ‘A’ pole and 575 Ma Callander pole. Meert and Van der Voo suggest that Sept-Iles ‘A’ is a result of later remagnetization, but it is also possible that the Callander pole is a result of remagnetization: the baked contact test for the Sept-Iles is more rigorous than for the Callander or Catoctin ‘A’. Meert and Van der Voo agreed with us about a 615 Ma age for the Long Range pole (Murthy et al. 1992; Kamo & Gower 1994). Any palaeogeographic model incorporating all of these three poles must include a low-latitude position at 615 Ma, high-latitude for 575 Ma and again low-latitude for 565 Ma. This requires unusually rapid continental motions. So, we prefer to consider one of two alternative models based either on the Callander pole, or on the Sept-Iles ‘A’ pole. Although we prefer the palaeogeographically and geodynamically simpler low-latitude model, we agree that there is not enough evidence yet to reject the other model. Meert’s and Van der Voo’s argument that Callander and Catoctin A poles match no known younger poles from Laurentia is strong. However, the time interval between 560 Ma and 530 Ma for Laurentia is poorly constrained by palaeomagnetism. Torsvik et al. (1996) reported about four poles with assigned ages of 550 Ma: Buckingham Flows (Dankers & LaPointe 1981), Long Range Dykes A (Murthy et al. 1992), Johnnie Rainstorm Formation (Van Alstine & Gillet 1979) and Double Mer Formation (Murthy et al. 1992). All these results show the low palaeolatitude. However, the only age constraint for the Buckingham Flows is the K–Ar determination of 573t32 Ma. Recent correlation of the Johnnie Formation with terminal Neoproterozoic strata in Australia suggest an age significantly older than 550 Ma (Wernicke & Hagadorn 2000; Abolins et al. Geophys. J. Int. (2001) 146, 871–873