Whitney et al. (2021) challenge our conclusions about rates of deformation and amount of uplift along the Cape Range, Western Australia, particularly the elevation constraints we place on the last interglacial shoreline along the northern half of Cape Range. They selectively focus almost entirely on the northern half of Cape Range, completely omitting our extensive analysis of the southern section, which provides the bulk of our paleo–sea-level interpretations. They also raise concerns about some of the nomenclature and methodology used. We thank them for the opportunity to clarify our results on the minor section of our paper they take issue to, and address their concerns below point by point.

We recognize that some of the terminology describing various structural and geomorphological features in the paper were used in a more general sense, which we considered appropriate for the journal. However, we appreciate some instances of ambiguity that more strictly defined specialist nomenclature could have avoided. As Whitney et al. (2021) mention, we defined relative deformation (or warping) and relative uplift (comparing northern, central, and southern Cape Range terrace elevations) as separate from absolute uplift, due to the complexities of having a truly independent datum to measure from. We tried to be specific when using the terms uplift or absolute uplift, and never claim that our method of calculating relative deformation is directly linked to absolute “tectonic uplift.” Nor do we ever use the specific term “tectonic uplift” in our paper, despite what is suggested in Whitney et al. (2021)—i.e., “tectonic uplift” appears 18 times in their comment. Furthermore, we never use the term “fold apex” as claimed by Whitney et al. (2021), but refer descriptively to the central apex of the anticline and apex of specific terrace features for their graphical representations where the highest elevation points are plotted, for both shorelines and Cape Range in general (as is clearly marked in fig. 4). The same is true for our use of the word limbs, with descriptive occurrences typically restricted to plots of specified shorelines or terraces on the northern and southern plunges of the fold on either side of their highest points (or culmination as is preferred by Whitney et al., 2021).

This section in Whitney et al. (2021) inaccurately represents what we present in our paper. As discussed above, we are clear and transparent about how our relative deformation, or warping rates, are calculated, with the primary goal of constraining the amount of deformation occurring at Cape Range since the Miocene. While we never use the term “tectonic uplift” in our paper, we do examine the ongoing relative uplift of the central Cape Range and constrain absolute uplift in the central and southern sections of Cape Range, as is evidenced in part by Miocene marine deposits reaching hundreds of meters in elevation above sea level (Riera et al., 2019, 2021). We know with high confidence that long-term uplift has been occurring for over 6.5 million years in the central and southern sections of Cape Range, primarily because the current elevation of the Muiron terrace in these sections is similar to, or higher than global mean sea level in an ice-free world (~66 m, Miller et al., 2020). This is further supported by evidence that late Miocene mean global sea level was most likely around 20 m above present levels (Miller et al., 2020), far lower than the current-day terrace elevations. Based on the premise of long-term uplift, with no evidence of subsidence occurring in this region of Cape Range, we are able to assign a maximum relative sea level for each of the terraces (but not a strict minimum, since we cannot know exactly how much uplift has occurred, as clearly specified in the paper). Whitney et al. (2021) incorrectly claim we have no “elevation datum,” or baseline from which to assess uplift over time, as there is an entire section in the discussion of Sandstrom et al. (2020) devoted to establishing mean sea level datums in the Pliocene and the implications for uplift of the Pleistocene shorelines. This is significant because we are able to constrain the degree of uplift for the younger shorelines more accurately, despite a relatively large uncertainty in Pliocene relative sea level. This method also avoids assumptions of last interglacial sea level that can form the basis of other works (e.g., Whitney and Hengesh, 2015).

There appears to be confusion on elevation measurements for individual transects that we wish to clarify. We measure the highest elevations of the same sea level features along each transect (e.g., the highest observed in situ corals). We then take into account the modern analogs of these features with respect to the tidal range (hereafter referred to as “indicative range”) as is reported in table S1 in Sandstrom et al. (2020), and use these relationships to determine paleo–sea level (Rovere et al., 2016). Any difference in elevation outside of the indicative range is therefore due to differential uplift along the anticline fold axis. We assume that the highest features recorded for each shoreline are recording a similar water depth (at least within the indicative range), and take the average of the three highest elevation differential global positioning system (dGPS) points on each transect (with uncertainty), where possible. This does mean the resulting average elevation is slightly less than the highest point, but our confidence in recording the true elevation, especially for marine limiting features (i.e., corals) is greatly increased. However, for full transparency, we have replotted our figure S4 (in what is now Fig. 1) showing the average elevation of Jurabi and Tantabiddi notches along with the highest elevation dGPS point for each transect.

Whitney et al. (2021) express apprehensions about the elevation differences of measured features on individual transects of the Jurabi terrace, as well as for elevation differences between neighboring transects for the Tantabiddi notches. We address their concerns below, but we would like to preface this by stating that some variation of sea-level proxy field data is to be expected due to a combination of variability in natural erosional/depositional/biological systems, imperfect preservation of sedimentary and fossil structures, as well as post-depositional processes that complicate data collection (Rovere et al., 2016; Hibbert et al., 2018). We attempted to mitigate these issues via the collection of high density, high precision dGPS points along numerous transects, and employed LOESS curves of best fit to help visualize the fundamental trends and reduce the impact of outliers. We believe this to be a more transparent approach than solely reporting the single highest elevation features.

What Whitney et al. (2021) call the “Jurabi Terrace” is actually the “Jurabi subtidal” of Sandstrom et al. (2020), representing the highest in situ coral and/or coralline algae (preferred marine limiting feature for this terrace, plotted in fig. 4b in Sandstrom et al., 2020). Each average elevation point is within 0.5 m of the highest recorded coral, and over 70% are within 0.25 m (see Fig. 1). The Jurabi beach sands are also replotted in Figure 1, with both the averages and individual highest points shown (two transects are over 0.25 m above the average, and the rest are below). Despite what Whitney et al. (2021) claim, every beach point agrees with the Jurabi subtidal data within the indicative range (table S1, Sandstrom et al., 2020) with the exception of two transects (labeled in Fig. 1), where the beach sands are slightly more than 0.37 m (our indicative range cutoff) below the Jurabi subtidal deposits. The Jurabi marine points are clast-supported matrix thought to be typically associated with shallow subtidal/intertidal environments (and include features such as articulated shells, oysters, etc.); however, the indicative range of these features is much larger and more uncertain than the other Jurabi terrace sea-level indicators. We therefore omitted them from our revised figure to better illustrate the relationship between the Jurabi subtidal and beach sediments. However, assuming an approximate indicative range of +1 msl (roughly in line with modern hightide) to −2 msl, we find that each Jurabi marine point is also in agreement with our Jurabi subtidal features. In summary, there is excellent agreement between the individual sea level features for measured Jurabi transects when accounting for the uncertainty associated with indicative range, and nothing like the 4–5 m discrepancies claimed by Whitney et al. (2021).

We also plotted the average and highest elevations for the upper and lower Tantabiddi notches/scarps (Fig. 1), with similar results as for the Jurabi terrace features. Only five points have individual highest elevations over 0.25 m above the average (0.71, 0.58, 0.34, 0.27, and 0.26 m above average), while the rest are below that value. Generally, there is good agreement between neighboring transects as well, as can be seen by the LOESS fit for both upper and lower notches/scarps. The sole discrepancy is at site TQS, where the upper Tantabiddi notch/scarp (only a single measurement) is higher than the nearby point by ~3 m and above the LOESS fit. The adjacent point to the south has been verified with LiDAR data; however, LiDAR coverage does not exist for site TQS. Analysis of the dGPS data shows it is conceivable that the scarp could be lower, somewhere between 10.57 m and 13.63 m (as we show in Fig. 1). However, we keep our original point at 13.6 m to maintain continuity with the other transects using the highest point(s) measured for the scarp or notch.

Whitney et al. (2021) prefer to treat all warping as cumulative when calculating uplift rates, such that with each successive terrace formation, all older terraces are “locked in” and experience the same amount of uplift as the youngest shoreline. In our paper, we chose to examine the deformation of terraces independent of each other, to assess how warping has changed per terrace based on the time since terrace formation. Although, as Whitney et al. (2021) point out, we fully recognize the successive nature of terrace formation and summative/incremental deformation that is responsible for the current deformation on older terraces, as had already been included in our figure (S7; Sandstrom et al., 2020). Broadly, nothing they say about warping of the anticline in this section contradicts what is stated in Sandstrom et al. (2020). Calculating the relative deformation rate individually (our method) relies on the age assigned to a specific terrace, while the cumulative calculation (Whitney et al., 2021, method) requires knowledge of ages from adjacent paleoshorelines, which are not always known, and which was exploited by Whitney et al. (2021) in their calculation of warping rates (Whitney et al., 2021, table 1), as we discuss below. Additionally, the shorelines are located at different distances from the fold axis (up to ~4 km horizontal distance perpendicular to the fold axis between the Muiron terrace and modern shoreline), which could potentially make the assumption “older terraces were merely passengers” (Whitney et al., 2021) invalid. Although this is probably a small effect, the different distances of the terraces from the fold axis makes it difficult to accurately collapse all terraces onto a single plot of progressive deformation due to complexities of variable deformation and dip on the limbs of a doubly plunging fold. Regardless, while we maintain that separate consideration of individual terraces remains a more transparent treatment of the data, in the following section we calculate the relative deformation rates through time using the cumulative method suggested by Whitney et al. (2021) and compare the results to our average terrace ages.

Whitney et al. (2021) object to our interpretation of two separate time-stratigraphic (upper and lower) Tantabiddi notches/scarps. They further state that because we provide no new age estimates on the notches, we should refrain from discussing the morpho-stratigraphy and interpretation of these topographically lower sea level features. We disagree, believing that we can make meaningful interpretations based on our new elevation and geomorphology data, in a similar manner to previous works that also lacked new age estimates (Whitney and Hengesh, 2015), or that were based on just one unpublished age estimate for the Jurabi Terrace (Clark et al., 2011). They also suggest there is no precedent for our interpretation of two separate time-stratigraphic features. This is not true. As we mention in our original paper, van de Graaff et al. (1976) discussed the presence of an older terrace a few meters above the last interglacial Tantabiddi terrace. The major difference in the interpretations between our paper and the comment appears to be whether what we refer to as the “upper Tantabiddi notch/scarp” formed during the last interglacial or is an older Pleistocene shoreline.

Whitney et al. (2021) maintain that separate elevations of the upper and lower Tantabiddi notches/scarps in the central and southern sections are due to paired low and high tide notches occurring at the same time during the last interglacial. In addition to all of the data outlined in Sandstrom et al. (2020) suggesting otherwise (such as the upper Tantabiddi scarp/notch more closely resembling the deformation and degree of lithification expressed in the Jurabi terrace than the other lower subtidal Tantabiddi deposits), this line of reasoning does not hold up under scrutiny. These types of paired notches form primarily in high wave energy environments (such as along the modern Quobba coast, 200 km to the south), whereas the entire western margin of Cape Range is protected by the Ningaloo reef in the modern setting, resulting in a single modern tidal notch (measured by Sandstrom et al., 2020, table S1). There is no reason to suppose that a similar fringing reef was not also present during the last interglacial, as is supported by the fossil record (van de Graaff et al., 1976; Sandstrom et al., 2020), which would have inhibited the high wave energy needed to create paired notches. Additionally, based on the elevation data in Figure 1 (and fig. S4 of Sandstrom et al., 2020), it is apparent that the upper Tantabiddi notch/scarp decreases in elevation northward from its highest level in the central part of Cape Range, while the lower Tantabiddi notch/scarp stays at a relatively consistent elevation. Eventually, both converge at the same elevation, with a single notch cut into the Jurabi terrace (Fig. 1). Systematic convergence of notches cannot be explained in a tidal range variation context over this distance (~28 km). Finally, based on the gathered elevation data (dGPS and LiDAR) at a number of spots along the Cape Range, we are able to further confirm that there is a distinct platform (often sheltered behind remnant marine isotope stage [MIS] 5e dunes) between the lower and upper Tantabiddi notches/scarps (see fig. 5 in Sandstrom et al., 2020).

Whitney et al. (2021) reanalyze our warping rates in their table 1, examining only the extreme minimum and maximum age constraints we place on the Tantabiddi features and solely focusing on warping rates in the northern to central section. We find these choices disingenuous, for two reasons. Firstly, the extremes of the presented age constraints do not reflect our best estimate age for the upper Tantabiddi notch. Secondly, we concentrate on constraining paleoshoreline elevations for the central to southern sections of the anticline in our paper (and specifically detail why the geometry in this region is more straightforward and robust for constraining relative sea levels in this area). Regardless, Whitney et al. (2021) appear to deliberately misinterpret our data by selecting the most extreme values that could conform to the age constraints provided, rather than most probable values, which create warping rates that are orders of magnitude larger than the older shoreline data would suggest. They purport that we infer that both the upper and lower Tantabiddi notches formed as two distinct features within the LIG (leading to a warping rate of 1.715 mm/yr, as highlighted by the gray box in their table 1). Not only do we never suggest this, but their own analysis from tables 1 and 2 suggests that the upper Tantabiddi notch cannot have formed during the last interglacial as the warping rate is around five times that of the older terraces. We recalculated the changes in warping rates of Cape Range with the same convention of Whitney et al. (2021) from their table 1, but use the average terrace ages we presented in Sandstrom et al. (2020). We use an estimated age of 0.410 Ma for the upper Tantabiddi terrace, consistent with the hard age constraints provided, and plot these data in Figure 2. The resulting warping rates are similar to those of figure 8 from Sandstrom et al. (2020), demonstrating how comparable these two methods are. While the deformation rate indicated by the upper Tantabiddi feature is still double any of the older warping rates calculated (perhaps because the age is actually older, or because the low signal-to-noise ratio for these younger shorelines translates small elevation discrepancies into large errors), it is far closer to the long-term average relative deformation rates of all the older shorelines. Assuming the lower Tantabiddi notch/scarp is MIS 5e results in warping rates similar to those of the older shorelines (Fig. 2).

We appreciate Whitney et al.'s (2021) acknowledgment of our contribution to the geochronology of the three older terraces and mapping of the region, which comprises the bulk of our paper. We also thank them for raising their concerns and allowing us the opportunity to clarify our methodology, nomenclature, and data on deformation rates in Cape Range. We maintain that our methods of calculating relative deformation or warping of these shorelines is correct, as is our interpretation of the upper and lower Tantabiddi notches/scarps representing two separate sea-level highstands, and show that our method for measuring shoreline features is accurate. We stress that our paper does not infer tectonic uplift rates, except in the southern section where we use Pliocene sea level estimates to help constrain the amount of uplift since the last interglacial. We value the focus and attention paid by the scientific community to this region, and hope future research endeavors continue to build upon the previous work to further our understanding of past climate and improve future sea-level rise predictions.

Science Editor: Brad S. Singer
Gold Open Access: This paper is published under the terms of the CC-BY license.