Reply to Hearty and Tormey: Use the scientific method to test geologic hypotheses, because rocks do not whisper

Hearty and Tormey (1) challenge our conclusions (2), incorrectly arguing that the megaboulders we discuss were shown to originate from the cliff bottom. A number of mischaracterizations are made by Hearty and Tormey (1) in their letter. First, we do not use a “tsunami wave model.” Second, we do not address the two other Bahamian landforms Hearty and Tormey (1) mention: their “superstorm” genesis interpretation [for which alternative hypotheses have been proposed (3, 4)] has no bearing on our (2) conclusions.

Hearty and Tormey’s (1) claim that the boulders have “fingerprints” based on “several physical criteria” and “data from multiple disciplines” is false. Only two mega-boulder “physical properties” were reported by Hearty (5): (i) cementation, which in Bahamian outcrops is spatially heterogenous and reflects exposure history, not ages or stratigraphy; and (ii) grain composition, based on “hand-lens and thin-section analysis” from which no data are provided, and from which the boulders were characterized as “generally oolitic/peloidal” (5). This grain composition describes every cliff unit at this location (6) and neither criteria can speak to megaboulder provenance.

The remaining argument for cliff-bottom provenance rests on whole-rock amino acid racemization (AAR) data (d-alloisoleucine/l-isoleucine ratio: A/I). Hearty’s publications indicate that cliff-top rocks at Glass Window Bridge have A/I values within the range of the megaboulders (A/I ranging from 0.60 to 0.74) (figure 5 in ref. 5; figure 4 in ref. 7). These data alone are sufficient to reject the notion that boulder A/I values are diagnostic for cliff-bottom units. Several cliff A/I ratios reported in figures 4 and 5 of ref. 5 (reproduced by Hearty and Tormey as figure 7 of ref. 8) are averages of samples taken throughout Eleuthera, yet are presented as though they represent a single stratigraphic sequence. In the Encyclopedia of Scientific Dating Methods, Murray-Wallace (9) warns against this practice: “results from one field area cannot be assumed to directly apply to another, even within a restricted geographical area.”

One of two subboulder AAR values used to argue the boulders sit upon younger units (5) was collected from an area that receives nearly daily wetting by sea spray and routine seawater inundation during storms, factors known to compromise AAR (9) values (by impacting temperature, moisture, and pH). Furthermore, these samples were collected immediately below an “entisol” (5), yet A/I samples were later rejected by Hearty and Kaufman (10) “on the basis of proximity to soil horizons.” These factors demonstrate that AAR-based inferences concerning boulder provenance are unsubstantiated and dubious.

Hearty and Tormey (1) question the apparent lack of modern analogs for the megaboulders. However, the authors (8) know that waves from Hurricane Andrew (1992) moved the Glass Window Bridge (463 tons, 12 m above sea level) off its foundation. Storms at this location are powerful enough to move massive elements located near the cliff edge, including numerous boulders smaller than the Cow and Bull. It has been widely observed that “where deposits sit atop a vertical cliff, clasts come from the upper part of the cliff, extracted and transported inland by the highest-reaching waves” (11).

Our conclusion that the Eleuthera megaboulders are not evidence for MIS 5e superstorms follows logically from the available data, and is supported by our analyses.