Page 8 - i1052-5173-32-6

Figure 3. (A) Schematic reconstruction of plate positions relative to Laurentia ca. 1100 Ma during formation of Rodinia. After the Elzeverian and
Shawinigan orogenies, but before the Grenville orogeny, spreading likely initiated between the major plates. Following failure of the Midcon-
tinent Rift (MCR), Amazonia shifted north along the margin before recolliding. (B) Similar reconstruction at ca. 560 Ma as Rodinia was breaking
up. Cuyania (Cu) block rifted off Laurentia, leaving the Southern Oklahoma Aulacogen (SOA) and Reelfoot Rift (RR) as failed arms. (C) Apparent
polar wander (APW) path of Laurentia, plotted in present-day coordinates, at 10-m.y. increments. Red cusp (1200–1000 Ma) is related to forma-
tion of the MCR, and blue cusp (700–500 Ma) is related to initial rifting of the SOA and RR. Path between these events plotted in gray.

opened as arms of this triple junction but and erosion and removal of low-density sedi- RR’s cross-sectional area is much smaller
ultimately failed (Fig. 3B). ments brought the volcanics closer to the sur- (160 km ). How these differences arose is
2
face, causing today’s gravity high. Without unclear. The volumes of igneous rocks pro-
Spatial Scale and Architecture this inversion, a positive anomaly would not duced in rifting can reflect two effects. The
The three rifts have similar spatial scales have developed. first is passive rifting in which extension due
and structures that seem to characterize We explored the hypothesis that inversion to far-field forces causes lithospheric thinning
failed rifts. Their central grabens, filled with is crucial for producing a positive gravity and inflow of hot asthenosphere, such that
volcanic and sedimentary rocks, are bounded anomaly using the SOA and RR. The SOA greater extension produces more melt (Koptev
by faults that presumably had normal fault experienced up to 15 km of inversion in the et al., 2015). The second, active rifting,
motion during extension. Despite structural late Paleozoic (Keller and Stephenson, involves an upwelling thermal plume, such
differences, all three rifts are ~60–80 km 2007). “Uninverting” the rift by re-burying that melt is generated by elevated mantle tem-
wide, suggesting that failed rifts are consis- the gabbroic fill 12 km below a sedimentary peratures beneath the lithosphere (Burov and
tent with observations that presently spread- basin eliminates the positive anomaly (Fig. Gerya, 2014). The relative roles of these and
ing rifts had initial widths controlled by 2E). Hence the SOA’s gravity high largely other possible rifting processes (King, 2007)
crustal thickness rather than the extension reflects the inversion. Conversely, because are extensively debated but remain unclear
history (Allemand and Brun, 1991). the RR did not experience significant inver- (Foulger, 2010). Both active and passive rift-
For the MCR and SOA, the rifting faults sion, its rift basin is buried beneath low- ing have been invoked to explain the volumes
were reactivated as reverse faults during sub- density sediments. Inverting the RR by 3 of volcanic rocks at rifted continental margins
sequent inversion. The SOA’s gravity high km and removing sediments overlying the (White and McKenzie, 1989; Richards et al.,
reflects structural inversion of basaltic and basin (Fig. 2F) produces a positive anomaly 1989; van Wijk et al., 2001). Gallahue et al.
gabbroic material in the Wichita Mountains, due to the high-density igneous rift fill being (2020) find evidence for both processes on
but significant amounts of rift-fill remain much nearer to the surface. continental margins, with passive rifting hav-
buried beneath the Anadarko Basin (Keller ing a stronger effect.
and Stephenson, 2007). Although the RR Igneous Rock Volumes A plume contribution for the MCR has
looks similar overall, it was not significantly There are interesting differences in the been inferred from petrologic and geochemi-
reactivated by later inversion. This left its volumes of rift volcanics. The MCR is ~3000 cal data (Nicholson et al., 1997; White, 1997;
rift-filling volcanics deeper in the subsur- km long and contains more than 2 million Davis et al., 2021), consistent with the enor-
face, causing the absence of a positive grav- km of buried igneous rocks, while the SOA mous volume of volcanic rocks making it a
3
ity anomaly. This effect is illustrated by a and RR are both roughly 1/10 the length of Large Igneous Province (Green, 1983; Stein
model showing the gravity anomaly at dif- the MCR and contain significantly less vol- et al., 2015). The large volume of MCR rocks
ferent stages in the MCR’s evolution (Fig. 4), canics. Although the SOA’s volcanic package also likely reflects Precambrian mantle tem-
derived from cross-section–balanced recon- produces a large positive gravity anomaly, peratures higher than today’s (Korenaga,
structions from GLIMPCE data (Stein et al., it contains only ~1/10 as much volcanics as 2013). The difference between the west and
2015). During rifting, dense volcanics near the MCR (Hanson et al., 2013). east arms likely reflects a difference in the
the surface would have caused a large posi- The differences appear in the cross sec- amount of extension during rifting (Merino
tive anomaly. Subsequent deposition of low- tions. Volcanics in MCR’s west and east arms et al., 2013; Elling et al., 2020). The smaller
density sediments and subsidence that have average cross-sectional areas of 1100 cross-sectional areas of volcanics in the SOA
depressed the volcanics would have caused a km and 680 km , the SOA has an average and RR probably do not require assuming
2
2
gravity low. Eventually, inversion of the rift cross-sectional area of 470 km , whereas the a plume. Hence, in our view, the simplest
2
8 GSA TODAY | June 2022