In the early 1960s, fieldwork and reading provided some clues. While there were no beach outcrops in the Port Stephens area where I was mapping Quaternary landforms, numerous drill holes supplied by the Hunter District Water Board in “sand beds” of Tomago and Fens (Tea Gardens) areas highlighted the existence of thick (up to 10m) organically cemented sand below what I was mapping as “Inner Barrier”. Harold Maze in 1942 had already written about the sand beds and “humus podsols” of the Newcastle district (Aust. Geog. 4, 107-112) so I was alert to a possible pedogenic origin of this material. It formed a hardpan from slightly above to below present sea level. This led me to question any previous suggestions that this barrier was Holocene in age. A paper published in 1965 (“Late Quaternary coastal morphology of the Port Stephens Myall Lakes area, NSW”, J. Roy. Soc. NSW, 98, 23-36) set out the case for a Pleistocene age of the Inner Barrier. Figure 3 showed what I then termed “sandrock” in a cross-section of the Fens Inner and Outer barriers.
A visit to the NSW north coast in 1961 was organised to follow the mapping of heavy mineral sand deposits by the Bureau of Mineral Resources (see Gardner, D., 1955, Bull. Bur. Min. Resour., 28, 1-103). These maps guided me to locations where “indurated sand” occurred below strandplain deposits and to outcrops on the beach. These black indurated sands were most striking in the vicinity of Jerusalem Creek south of Evans Head. Sedimentary structures in the outcrops suggested beach deposition containing seams of heavy mineral mined later at sites behind the beach such as “Macauley’s Lead”. What I was seeing here reminded me of Lake Cathie but on a bigger scale.
So just what is this “indurated sand”, where does it occur and how was it formed? Within the siliceous sand province of eastern Australia there occurs soils with hardened organic bands (sometimes with iron oxides). Recently Stephen Gale and colleagues in their studies of aeolian sands of the Botany Basin have summarised much of the literature and nomenclature around these materials ( see Gale et al., 2017 Aust. Geog., 49, 291-316; and Gale and Wales, 2022, Geomorphology , 405, 108175). In the 2017 paper they describe the complex nature of variably indurated layers found in the quartz sands of the area under the local name of “Waterloo Rock”. The layers have developed in at least two distinct contexts: one as a B horizon of a soil profile; and two as organic deposition close to the water table.
There were three further encounters that have helped me better understand the origin of indurated “sandrock” (now commonly referred to as “coffee rock”). The first follows from mapping and drilling undertaken in Horry County, South Carolina, where I did fieldwork in the mid-1960s. Again there existed organic indurated sands as hardpans within Pleistocene strandplain deposits. I discussed these in a short note on “Humate and coastal geomorphology” in 1967 (Coastal Studies Bulletin, LSU, No.1, 15-17). Here I used the term “humate” following the work in Florida of Swanson and Palacas (1965, US Geol. Surv. Bull., 1214-B, 1-29). They were referring to gel-like humic substances in sands that were carried in colloidal suspension or solution through void spaces to accumulate at the water table by precipitation or flocculation. The humic substances are derived by the leaching of decaying plant material. In this note I made comparisons with what I knew from eastern Australian making specific reference to the work of Ted Coaldrake in southern Queensland (Coaldrake, J.E. 1955, Aust., J. Sci., 17, 132-133; 1962, Proc. Roy. Soc. Qld., 72, 101-116). I was convinced that what we were looking at generically was a process linked to water table oscillations; in this respect they constituted “giant ground-water podzols”.
The next encounter was more indirect but still relevant to unravelling a controversy around late Quaternary sea level history. During the 70s marine isotope and uranium series dating were coming together to define stages of Quaternary sea-level change. John Chappell at ANU was party to this incredible work. But there was a problem. From various sources around the world radiocarbon dates on sea-level indicators were being published to define an “interstadial” high sea level close to present 30000 to 40000 C14 years ago. One site that was yielding such dates was Jerusalem Creek. Trevor Langford-Smith had sampled a log of wood within the beach deposits encased in coffee rock. I visited the site with Trevor and debated the issue. I knew that shells from below the indurated sand at Tomago were aged beyond the range of radiocarbon dating. I was encouraged by John and others to review the problem (published in 1973 in Progress in Geography, 5, 170-246). What was bothering me was whether there was any contamination by the humic colloids of the log. This question took me into the brave new world of organic chemistry.
The final encounter involved working with the team from the ANU Radiocarbon Laboratory on dating different humic acid fractions of samples from a drill hole within the Tomago Inner Barrier near Newcastle. Details of this work are to be found in the monograph on the Quaternary of the region (Thom et al., 1992, ANU Press). We termed the material that was fractionated and dated “humate-impregnated sand”. It was 7 m in thickness extending below sea level. We found an age range of 9000 years towards the surface going to 25000 years at the base (humic acids slightly older than fulvic acids at a given depth). These results tested two hypotheses, one that any wood sample could be contaminated by younger colloidal humic acid fractions; and two that any model of humate impregnation to form coffee rock involved downward percolation of organic-rich waters over time to a water table (below the water table there is no leaching allowing preservation of shells which C14 dated as background). The model required the water table position to progressively rise from a basal level fixed when sea level was lower during the Last Glacial (see Fig. 4-18 in Thom et al., 1992, p. 140). This “pulling up by the bootstraps model” links quite comfortably with the now established age of the Tomago Inner Barrier strandplain as Last Interglacial (MIS5e).
Coffee rock (by whatever name) is a major feature of coastal lands of eastern Australia within both beach and dune deposits especially those of Pleistocene age. It a characteristic of silicious sandy environments in contrast to aeolianite and beachrock found in places rich in carbonate sand. Where it outcrops along a beach, or even a coastal creek (e.g. near Wooli), it provides a degree of resistance to shoreline recession. It can vary in hardness. Exposed eroded surfaces frequently display signs of differential erosion. Why it sometimes appears to harden on exposure is difficult to understand., but on receding shores it can remain intact and occur offshore as submerged reefs (e.g. at Old Bar). Recent work by the Department of Environment, Science and Innovation in Queensland highlights “special values” of coffee rock as habitats.
Back to Lake Cathie. Two years ago the NSW Coastal Council was tasked by the Minister for Local Government to provide advice on a possible erosion threat to the road and properties at Illaroo Road, Lake Cathie, and whether a sea wall was needed. The backshore consisted of coffee rock that was periodically impacted by storm waves. Here I was some six decades on from first meeting the black indurated sand at this place contemplating future recession rate projections with colleagues. The rock appears to be doing its protective job but for how long?
Words by Prof Bruce Thom. Please respect the author’s thoughts and reference appropriately: (c) ACS, 2024. For correspondence about this blog post please email email@example.com