DEVONIAN REEF COMPLEXES OF THE CANNING BASIN, WESTERN AUSTRALIA: A HISTORICAL REVIEW
- File PDFChapter PDF
Phillip E. Playford, Roger M. Hocking, Anthony E. Cockbain, 2017. "DEVONIAN REEF COMPLEXES OF THE CANNING BASIN, WESTERN AUSTRALIA: A HISTORICAL REVIEW", NEWADVANCES IN DEVONIAN CARBONATES: OUTCROP ANALOGS, RESERVOIRS AND CHRONOSTRATIGRAPHY, Ted E. Playton, Charles Kerans, John A.W. Weissenberger
Download citation file:
Devonian reef complexes are spectacularly exposed in a series of limestone ranges along the northern margin of the Canning Basin in Western Australia and have become known as “The Devonian Great Barrier Reef.” The geological literature on these rocks dates back to 1884, and systematic research on them began during the late 1930s. Since then, many individuals and organizations have progressively increased knowledge of the stratigraphy and paleontology of the reef complexes, although one study concluded that they are products of “dynamic metamorphism.” Comprehensive research by the Geological Survey of Western Australia and its coworkers culminated in 2009 with the publication of a detailed account of the surface geology of the reef complexes and their associated terrigenous conglomerates. This article presents an overview of the research into the reef complexes, focusing on the key milestones and developments in knowledge and concepts.
INTRODUCTION: AN OVERVIEW OF THE REEF COMPLEXES
In order to better appreciate the evolution of our understanding of these reef complexes we first summarize our current knowledge of them and then discuss the successive contributions made by various workers, followed by a critique of the “metamorphic model” and a retrospective of over 130 years of work on these rocks.
Middle and Upper Devonian (Givetian, Frasnian, and Famennian) reef complexes are spectacularly exposed on the Lennard Shelf, along the northern margin of the Canning Basin (Western Australia), and are widely acknowledged as being among the best-exposed examples in the world of an ancient barrier reef system (Figs. 1,2). They form a belt of rugged limestone ranges, approximately 350 km long and up to 50 km wide, that is commonly known as “The Devonian Great Barrier Reef”. The reef complexes form a northwest-trending belt composed of fringing reefs, atolls, and banks that grew along the mountainous mainland shore of the Kimberley block and around rugged islands of Proterozoic igneous and metamorphic rocks. One reef complex grew on a fault block of Ordovician dolomite and shale. The maximum thickness of the Devonian rocks is estimated to be at least 2500 m.
In some areas the reef complexes are cut by normal faults, some of which moved during the Devonian, with associated tilting and folding, but over large areas the Devonian rocks remain almost undeformed. Contemporaneous terrigenous conglomerates interfinger with, or pass through, the reef complexes. These conglomerates were derived from the scarps of active faults cutting the Precambrian basement to the north. Movement along some faults continued during the Carboniferous, but since then there has been little or no faulting in the area.
Three main facies are recognized in the reef complexes: platform, marginal-slope, and basin facies. The reefal platforms stood tens to hundreds of meters above the adjacent sea floor and were constructed by shallow-water organisms, especially stromatoporoids, corals, and microbes. Many platforms were rimmed by rigid wave-resistant reefs. The platform facies is subdivided into reef-margin, reef-flat, pinnacle reef, and back-reef subfacies. Where no reef is developed around a platform margin, the platform is regarded as a bank and its deposits as bank subfacies.
The platform deposits were laid down essentially horizontally in shallow subtidal, intertidal, and supratidal environments. Reef-margin and reef-flat deposits formed mainly in shallow water depths, but in places a reef grew in water up to a few tens of meters deep. Back-reef areas ranged from supratidal to subtidal, with estimated water depths of up to 10 m. Cyclicity is evident in many of the back-reef deposits.
Marginal-slope deposits were laid down on slopes in front of the platforms, descending to water depths up to several hundreds of meters deep. The marginal-slope facies in front of each reefal platform is subdivided into reefal-slope and fore-reef subfacies or fore-bank facies where the platform was a bank.
Reef-margin and reef-flat boundstones and back-reef biostromes were built by microbes, stromatoporoids, and corals during the late Givetian and early Frasnian, microbes and stromatoporoids during the late Frasnian, and microbes alone in the Famennian. The reefal-slope subfacies consist of microbial boundstones that accreted at the tops of the marginal slopes. The reefal-slope deposits show depositional dips, ranging from nearly vertical to about 40°, and pass downwards into fore-reef subfacies.
Fore-reef deposits consist largely of platform-derived debris and include debris flows and isolated allochthonous blocks of reef together with indigenous fossil organisms and terrigenous clastic material. Depositional dips in the fore-reef subfacies decline progressively from about 40° at the top of a slope to a few degrees at the foot, where the fore-reef subfacies interfingers with basin facies. Fore-bank deposits generally lack steep depositional dips and interfinger directly with bank deposits at the top of the slope and with basin facies at the base.
The basin facies was laid down essentially horizontally in water depths ranging from a few tens to several hundreds of meters and consists largely of calcareous shale, siltstone, and sandstone, with some interbedded turbidites and debris-flow limestones. Most basin deposits have undergone major postburial mechanical compaction (up to about 75%).
The reef complexes range in age from Middle Devonian (late Givetian) to Late Devonian (Frasnian and Famennian). Most exposed reefs are Frasnian and Famennian in age. The most precise dating is based on conodonts and ammonoids in basin and marginal-slope deposits. Conodonts and ammonoids are very rare in platform deposits. Two second-order sequences are recognized in the reef complexes: the Givetian-Frasnian Pillara Sequence and the Famennian Nullara Sequence. The boundary between these sequences is an unconformity in platform and upper marginal-slope deposits, grading to a conformity in deeper marginal-slope and basin deposits. The fall in sea level that caused this unconformity is estimated to have been about 50 m. The Frasnian-Famennian boundary marks the culmination of a global mass extinction of metazoan organisms that began during the late Frasnian. Microbes survived the mass extinction virtually unscathed. Among those microbes, Renalcis is especially prominent as a reef builder in Frasnian and Famennian platforms, but nonskeletal microbes were also important as reef builders.
Deep-water stromatolites are conspicuous features of some marginal-slope deposits, above and just below the Frasnian-Famennian boundary. They may have thrived at that time because the extinction event removed metazoans that would otherwise have consumed the stromatolite-building microbes.
The rigid early-cemented reef-margin and reef-flat limestones were subjected to fissuring in response to earthquake shaking, slippage along underlying marginal-slope deposits, and differential compaction of underlying basin deposits over basement topography. The fissures were filled with sediment, calcite cement, and organic growths, forming networks of neptunian dykes. Masses of terrigenous conglomerate interfinger with and extend through the reef complexes at several locations along the outcrop belt. They interfinger with platform, marginal-slope, and basin deposits and were laid down as alluvial-fan, fan-delta, and submarine-fan deposits in front of the scarps of active faults. Large volumes of sand and mud poured into basins adjoining the conglomerate bodies, so that the resulting basin deposits are largely terrigenous.
The area was subjected to glaciation by continental ice sheets during the Late Carboniferous and Early Permian. The erosive action of the ice sheets and associated subglacial water had profound effects on the Devonian rocks. The tops of the limestone ranges were planed off by “dirty” ice and extensively karstified by the corrosive action of subglacial water under high pressures and subzero temperatures. Major cave systems formed in the limestones at that time.
Economic deposits of zinc and lead sulfides have been mined along the reef belt, mainly in the southeastern part of the belt. These deposits are thought to have been developed in the Devonian limestones by hot fluids expelled from shales deep in the Fitzroy Trough. They follow faults and hydrothermal caverns in the limestones. The age of this epigenetic mineralization is Early Carboniferous (Tournaisian).
Small oil fields have been located in late Famennian reef limestone and overlying deposits in the subsurface of the northwestern Lennard Shelf. The Famennian reef margin has been well defined in this area through conventional seismic surveys. Although Frasnian reef complexes are known from drilling to occur below the Famennian carbonate rocks, their detailed distribution cannot be delineated by conventional seismic surveys.
The progressive understanding of the reef complexes, as expressed in various publications and reports, is the main focus of this article. For a comprehensive bibliography refer to Playford et al. (2009).
EVOLUTION OF OUR UNDERSTANDING OF THE REEF COMPLEXES
Early Geological Investigations
The first geologist to examine the rocks of the reef complexes was E.T. Hardman, who, as a temporary Government Geologist, accompanied John Forrest’s expedition to the area in 1883 (Hardman 1884). He examined the Napier Range at Windjana Gorge (“Devil’s Pass,” Fig. 3), Geikie Gorge, Mt Pierre, and adjoining areas. Hardman did not recognize these rocks as reefal deposits and concluded that they were of Carboniferous age. Examination of his fossil collections soon showed them to be Devonian (Foord 1890, Hinde 1890, Nicholson 1890).
While working for the Freney Kimberley Oil Company during the 1920s and 1930s, Arthur Wade was the first to recognize these limestones as reef deposits (Wade 1924), later describing them as an “ancient barrier reef” (Wade 1936).
Pioneering Research: Curt Teichert, then a staff member of the University of Western Australia, studied the Devonian reefs during the late 1930s and early 1940s in association with the geologists of Caltex (Australia) Oil Development Pty Ltd., who were then assessing the oil prospects of the Canning Basin. Teichert examined the Devonian paleontology and biostratigraphy and published a series of articles on what he termed the “Great Devonian Barrier Reef” (Teichert 1949). He was the first to recognize the facies equivalence of various parts of the complexes and their associated conglomerates (Fig. 4). Although he was able to spend relatively little time in the field, Teichert laid firm foundations for subsequent, more-detailed studies of the reef complexes. In recognition of those important contributions, a well-preserved platform atoll in the Old Bohemia area has been named “Teichert Hills.” This atoll is especially noteworthy for chute-and-buttress (or spur-and-groove) structures, which are well exposed along its southwest margin (Fig. 5).
Regional Mapping by the Bureau of Mineral Resources: From 1948 to 1952 the Commonwealth Bureau of Mineral Resources mapped the full extent of the Devonian outcrops for the first time as part of a regional geological survey of the northern Canning Basin (then referred to as the “Fitzroy Basin”). In 1953 the manuscript of a bulletin on the geology of this area, together with relevant field books, was destroyed in a fire at the Bureau’s offices in Canberra. The bulletin (Guppy et al. 1958) was subsequently rewritten by the same authors, all but one of whom had moved to West Australian Petroleum Pty Ltd. (WAPET). They defined many rock units in the reef complexes and recognized that the steep dips in parts of those complexes are depositional. However, they did not recognize the equivalence of various facies, and they mistakenly concluded that the Pillara Limestone is entirely Middle Devonian (Givetian) in age and is overlain unconformably by Upper Devonian strata with steep depositional dips (Fig. 6).
Early Systematic Mapping
WAPET recognized that if these Devonian reef complexes extend into the subsurface, they could have a high potential for petroleum, because similar Devonian reefs in Alberta were known to host large oil reserves. The company had access to the Guppy et al. (1958) manuscript two years before its publication and determined that a more detailed study of the reef complexes was warranted. Consequently, a number of detailed studies were undertaken, starting in 1956 with a study of the Oscar Range reef complex (Smith et al. 1957).
Oscar Range: In this study Smith and Williams were responsible for the mapping, while Playford made periodic visits to the field, studied the petrology of the rocks, and was responsible for becoming familiar with the literature on ancient reef complexes. Perhaps the most significant of those publications was that of King (1942) on the Permian reef complex of West Texas and southeastern New Mexico, in which King showed that distinct facies could be recognized in the reefal deposits (Fig. 7). Playford considered that comparable facies might also be represented in the Oscar Range reef complex, and he and his fellow workers sought to verify this from field relationships. Indeed, it was confirmed: three basic facies—reef, back-reef, and fore-reef—could be recognized (Smith et al. 1957; Figs. 8, 9). It was also recognized that the steep dips seen in fore-reef deposits are largely depositional and that the Pillara Limestone consists of the oldest back-reef and reef deposits, equivalent to older parts of the fore-reef facies. The stratigraphic names these authors introduced were informal (Figs. 8, 9) and were not adopted in later studies of the reef complexes.
Another important outcome of the study by Smith et al. was the recognition of the major role played by microbes in constructing the reefs. Some of those microbes, first observed in thin sections, were initially termed “ghost algae” and were later recognized as the microbe Renalcis (Playford and Lowry 1966, Playford 1967, Wray 1967). It would also be shown that this microbe was an important constituent of the Canadian Devonian reef complexes (Playford 1969).
Windjana Gorge: The focus of WAPET work moved to the Upper Devonian reef complex at Windjana Gorge in 1958 (Playford and Johnstone 1959, Playford 1961; Figs. 10–13). A notable outcome of that work was recognition of the spectacular exposure in the gorge that would later become known as “The Classic Face” (Fig. 13). Boulder conglomerate, derived from Precambrian rocks of the Kimberley block, was shown to interfinger with the various facies of the reef complex that are exposed at and near the east entrance to the gorge (Fig. 11).
In discussions that followed a presentation on Windjana Gorge at the Geological Society of America’s annual meeting in 1961 (Playford 1961), Lloyd Pray suggested that the isolated “bioherms” that Playford had shown in his slides might instead be allochthonous blocks of reef limestone. When Playford and Lowry began field work on the reef complexes in 1962 it soon became clear that this deduction was correct—those blocks in the fore-reef deposits are indeed allochthonous, although it was later recognized that some are capped by deep-water microbial bioherms.
Early Geological Mapping by the Geological Survey of Western Australia: Systematic mapping and interpretation of the reef complexes was conducted during 1962 and 1963 by the Geological Survey of Western Australia (GSWA), the results being published in Bulletin 118 (Playford and Lowry 1966). Among the main outcomes of that study were recognition of classic atolls and pinnacle reefs in the eastern part of the outcrop area (Figs. 14, 15) and of the importance of contemporary tectonism in controlling development of the reef complexes. Fracturing of early-cemented limestones, resulting from that tectonism, was shown to have developed networks of neptunian dykes (Fig. 16). Neptunian fracturing and collapse also gave rise to megabreccia debris flows and isolated allochthonous blocks that came to rest on marginal slopes in front of the reefal platforms (Figs. 17, 18).
Detailed Studies of the Reef Complexes
With the broad outline of the reef complexes now known, work concentrated on a number of detailed studies of features of particular interest and significance.
Deep-Water Stromatolites: The GSWA, in association with the Commonwealth Bureau of Mineral Resources, conducted detailed studies of the reef complexes in the Bugle Gap area in 1968. One important outcome of that work was recognition of the use of geopetal structures to quantify depositional dips and to deduce paleobathymetry (Figs. 19–21). It was shown that some stromatolites must have grown on marginal slopes in water depths of at least 35 m (Fig. 18) and probably more than 100 m (Playford and Cockbain 1969). That conclusion was contrary to the commonly held belief that stromatolites are exclusively intertidal in origin (Logan 1961).
Cyclicity:Read (1973a, 1973b) was the first to recognize and document cyclicity in the Frasnian platform limestones. Shallowing-upward cycles in back-reef and bank subfacies are now deduced to be eustatic in origin (Brownlaw et al. 1996). Subsequently, Hocking and Playford (2001) showed that two scales of cyclicity (meters and tens of meters) are present, and Playford et al. (2009; Fig. 22) showed that Milankovitch cyclicity can also be recognized in other facies of the reef complexes.
Deep-Water Bioherms:Playford et al. (1976) described the presence, in the northeastern Oscar Range, of large Devonian microbial and receptaculitid bioherms, named the Elimberrie bioherms (Fig. 23). These grew in deep water over drowned pinnacle reefs and may be unique in the world.
“Metamorphic Model”: A special publication of the Geological Society of Australia described the Devonian limestones as products of dynamic metamorphism (Logan and Semeniuk 1976) (Fig. 6). Those authors asserted that the limestones do not form reef complexes but instead are the products of dynamic metamorphism, associated with intensive shear faulting and with metamorphic grades as high as greenschist facies. Similar claims were promulgated by Logan (1984) and Logan et al. (1994). However, those conclusions were not accepted by other authors, all of whom agree that the limestones represent unmetamorphosed reef complexes and that the supposed shear faults do not exist (see below for a fuller critique of this model).
Retreating, Back-Stepping, and Advancing Platforms:Playford (1980) published an article on “The Devonian Great Barrier Reef of the Canning Basin, Western Australia,” following a 1978 AAPG Distinguished Lecturer tour of the United States and Canada. A notable feature of this article was the recognition of Givetian-Frasnian retreating and back-stepping reefal platforms, succeeded by Famennian advancing platforms.
Facies Nomenclature and the Frasnian-Famennian Mass Extinction:Playford (1984) presented an updated facies nomenclature for the reef complexes (Fig. 24) and also discussed the mass extinction at the Frasnian-Famennian (F/F) boundary. That mass extinction resulted in the loss of many marine species, so that the reef-building stromatoporoids and corals of the Frasnian (Fig. 25) were replaced by microbes, almost alone as reef builders, during the Famennian (Fig. 26). Marked changes also occurred at that boundary in the conodont and ammonoid faunas of basin and marginal-slope deposits. The F/F boundary is unconformable in the reef, reefal-slope, and back-reef deposits and conformable in the deeper water fore-reef and basin facies (Figs. 27, 28).
Petrology of the Reef Complexes:Kerans (1985) made the first detailed studies of the petrology of the reef complexes, emphasizing the importance of marine cementation in early diagenesis. The strongest early cementation, and concomitant destruction of porosity, occurred in reef-margin, reef-flat, and reefal-slope deposits (Figs. 29–31).
Paleomagnetism:Hurley and Van der Voo (1987) examined the paleomagnetic stratigraphy of limestones from seven sites (89 samples) in the reef complexes, determining that these rocks were deposited at 15°S latitude, which is consistent with reef growth in this area during the Late Devonian.
Further paleomagnetic studies of the Late Devonian reef complexes in the Canning Basin were made by Li et al. (1993) as part of their research on polar wandering in Gondwana. Their results for the Late Devonian in the Canning Basin confirmed those of Hurley and Van der Voo (1987).
Hansma et al. (2015) reported on a detailed magnetostratigraphic and biostratigraphic study, at decimeter to meter scales, of samples from middle Frasnian to Famennian marginal-slope limestones at the Oscar Range (754 samples) and the Horse Spring Range (295 samples). This was part of a wider multifaceted study of the chronostratigraphy of the reef complexes (see Playton et al. 2016). They concluded that those rocks were deposited at 9.9°S latitude during the Late Devonian. Multiple magnetic reversals were observed, and magnetic polarity stratigraphy enabled the correlation of 12 magnetostratigraphic “packages” between the two locations.
Sequence Stratigraphy:Kennard et al. (1992) were the first to apply principles of sequence stratigraphy to the reef complexes. They adopted the “Exxon paradigm,” whereby eustatic fluctuations in sea level are claimed to drive reciprocal sedimentation of highstand, transgressive, and lowstand “system tracts.” Their reciprocal model interpreted deposition of terrigenous conglomerates as lowstand deposits and of reefal platforms as highstand deposits, with thinner intervening transgressive deposits (Figs. 6, 32). However, the field studies of Playford et al. (2009) showed that terrigenous conglomerates exposed in the area interfinger with platform, marginal-slope, and basin facies and persist over many eustatic cycles. They are probably derived from the scarps of active faults. Playford et al. (2009) did not recognize any of the transgressive deposits suggested by Kennard et al. (1992) and regarded the outcropping reef complexes as entirely highstand in origin.
The similarities between growth patterns of the Devonian reef complexes in Alberta and the Canning Basin strongly suggest that long-term global sea-level changes have been important in controlling development of these widely separated reef complexes (Playford et al. 2009, fig. 85), with tectonism important on a shorter time frame.
Deep-Water Stromatolite Mounds and Sulfide Mineralization: Deep-water stromatolite mounds associated with barite mineralization and cut by iron-sulfide veins were formed as exhalative deposits over cool-water seepages on the basin floors (Playford and Wallace 2001; Fig. 33). Those exhalative deposits resulted from compaction-driven fluids expelled from anoxic muds of the basin facies. In addition to nourishing deep-water stromatolites, these fluids gave rise to associated barite and sulfide mineralization. A wide variety of other stromatolites grew on shallow reefal platforms and adjoining marginal slopes, where they are associated with open-marine benthic faunas, whereas the exhalative stromatolites grew on and below the muddy floors of deep-water basins, without any associated benthic faunas (Fig. 34).
Studies by The University of Western Australia: Commencing in the mid- to late 1990s, A.D. George and coworkers from the University of Western Australia and elsewhere published a series of articles, generally focusing on tectono-stratigraphic and sequence-stratigraphic aspects of the evolution of the reef complexes, in particular dealing with possible controls by syndepositional faulting (George et al. 2002, 2006, 2009a, 2009b, 2013). Although it has long been known that there is clear evidence for such faulting, we do not concur with many of their other observations and conclusions (e.g., see discussion in George et al. 2006, Playford and Hocking 2006, and Playford et al. 2009). However, a detailed critique of that work is beyond the scope of this article.
Permian Glaciation:Playford (2002) discussed the important role of the Permian glaciation in leveling the reef complexes, below thick ice sheets that moved from south to north (Fig. 35). Subglacial water below those ice sheets resulted in extensive networks of Nye channels and subglacial karst (Figs. 36, 37). The present exposure surfaces on the complexes are largely a result of that glaciation.
Paleontology: The Devonian paleontology of the Canning Basin is renowned worldwide, but a review of the work on individual groups needs a separate and lengthy article. The most recent publications on important groups have been by Wray (1967) on microbes; Cockbain (1984) on stromatoporoids; Won (1997) on radiolarians; Jell and Jell (1999) on crinoids; Brownlaw and Jell (2008) on corals; Klapper (2009) on conodonts; Becker and House (2009) on ammonoids; G. Playford (2009) on palynology; Feist and McNamara (2013) on trilobites; and Trinajstic et al. (2014) on fish (the beautifully preserved fossils extracted from concretions in the Gogo Formation are world famous).
GSWA Bulletin 145: A detailed synthesis of the results of mapping and associated research on the Devonian Great Barrier reef by GSWA and its collaborators was published by Playford et al. (2009) as Bulletin 145 of the Geological Survey (freely available as PDF from GSWA website). This publication is based mainly on remapping and detailed stratigraphic studies of the reef complexes conducted during the 1990s, with associated paleontological research. It refined the presentation of the facies and stratigraphy and presents an event-based sequence stratigraphy of the reef complexes. The bulletin includes a comprehensive guide to important field localities and more than 530 color photos and diagrams.
Figure 38 shows the various subdivisions recognized by these authors in the reef complexes, and Figure 39 shows the lithostratigraphy, sequence stratigraphy, and event stratigraphy responsible for back-stepping and partial drowning of the reefal platforms. Siliciclastic conglomerates are shown to be tectonically driven and synchronous with the reef complexes. Tectonic and eustatic components are both thought to have influenced cyclicity and facies developments.
Chronostratigraphic Studies: The Canning Basin Chronostratigraphy Project (Playton et al. 2016) is a collaborative project between industry, several universities, and the GSWA, building on the work of Playford et al. (2009). Outcrops and cores were logged and sampled at submeter scale for magnetostratigraphy, stable-isotope stratigraphy, conodont and fish biostratigraphy, biomarker geochemistry, and elemental chemostratigraphy. The data set encompasses 4 km of measured described stratigraphy and 6800 samples of platform, reef, marginal-slope, and basin deposits. The samples yielded independent signals in the rock record that, when integrated with correlation constraints, generated a multifaceted, regional shelf-to-basin chronostratigraphy across the Lennard Shelf.
Geochemical Studies:George et al. (2014) and Hillbun et al. (2015) conducted geochemical studies across the F/F boundary and surrounding precursor and recovery intervals in marginal-slope facies of the reef complexes. The work of Hillbun et al. is one facet of the Chronostratigraphy Project and identified global isotope excursions (Lower and Upper Kellwasser Events) that were used for correlation control around the data set. George et al. also identified the Kellwasser Events in a core study but showed differences in the timing of excursions when compared to Hillbun et al. Further geochemical work in key localities, such as in Windjana Gorge itself, where the F/F boundary is constrained to a single bedding plane, is needed to understand the discrepancies between the two studies; however, both reveal for the first time the occurrence of the Kellwasser excursions in the Southern Hemisphere and their preserved signal despite the lack of organic-rich facies.
THE “METAMORPHIC MODEL” OF LOGAN AND SEMENIUK
Logan and Semeniuk (1976, p. 1) considered that “The majority of Devonian carbonate rocks cropping out along the northern margin of the Canning Basin are metamorphic products which have resulted from chemical solutions acting in a stressed environment.” Their “metamorphic processes” include “pressure solution, fracture, dolomitization, solution, emplacement, and recrystallization.” While using the stratigraphic terminology of Playford and Lowry (1966) they emphasized that “such usage does not, however, imply our acceptance of the ‘reef’ interpretation” (Logan and Semeniuk 1976, p. 9). Indeed, their claims are incompatible with that interpretation. Logan (1984) and Logan et al. (1994) further developed this metamorphic model, but their interpretations have not been accepted by any other authors, nor have they been applied to deposits elsewhere in the world.
In reviewing the Playford et al. (2009) bulletin, Martindale (2009) and James (2010) inferred that the dynamic-metamorphism interpretation should have been discussed in more detail. Indeed, there have been no detailed discussions of Logan and Semeniuk’s interpretations by other authors, all of whom have essentially dismissed them out of hand, presumably because they are so obviously wrong. Consequently, in this article we discuss some of Logan and Semeniuk’s key points; a more detailed critique is beyond the scope of this article. Page numbers quoted in the following refer to those in Logan and Semeniuk (1976).
Central to Logan and Semeniuk’s metamorphic model is the “iden,” which they defined as a “body that behaves as a statistically homogeneous entity to physical or chemical stimuli … it is also a finitely extended body independent of scale, geometry and composition” (p. 6); and “an iden can range in size from the smallest particle to the massif of the structural geologist and beyond; in shape from regular geometric to irregular … in petrographic terms it may be sedimentary, metamorphic or igneous” (p. 8). Thus, the iden of Logan and Semeniuk can be any object, of any size, composition, and origin, and can range in size from a minute grain of silt to a continental massif. They also defined many other new terms that they considered to have “wide applicability to carbonate rocks.” However, no subsequent authors have accepted their terms or made interpretations using their concepts.
Logan and Semeniuk’s figure 83, copied here as Fig. 40 and entitled “Paths by which polyidenic skeletal-boundstone pods developed,” shows their interpretation of the origin of megabreccias, supposedly through combined shear stress and stylolitization. They did not discuss the interpretations of any other authors who have concluded that the megabreccias in these rocks are accumulations of boulders derived from the collapse of reef margins and slides on fore-reef slopes.
An example of their thinking on the supposed origin of megabreccias is that “megabreccias are units produced by shear fracture, as indicated by discordant relationships, drag-folding of adjacent stylobeds, and rotated fragments, many of which are rocks with metamorphic structures and fabrics ” (p. 110). Further, they claim that megabreccias “have an origin similar to stylobreccia, i.e. the fracture and pressure solution along dislocation zones and the only real difference between theme [sic] is scale: megabreccia sheets developed along major faults which split and coalesced, at times isolating lenses of host rock in the breccia,”and “brecciation probably occurred during periods of rapid relief” [sic].
Many statements made in the text are similarly obscure, with extensive use of newly introduced jargon, leading to claims that are difficult or impossible to follow. For example: “Massive structureless dolo-interlayered calcareous idens occurs in the lower to middle Napier Formation where dolomite occurs in a sheetlike concordant body up dolomitic stylolaminites are interlayered with sheets of quartzose stylolaminite and limestone” [sic] (p. 59).
Logan and Semeniuk (p. 9) considered that horizontal to gently inclined blocks of “Pillara Formation (i.e. Pillara Limestone) meet a steeply dipping homocline of Napier, Virgin Hills or Sadler Formations”, the contact between the so-called “block” and “homocline” being termed the “abutment zone” (Fig. 41). That zone is claimed to be “a melange of material” filling “a zone of dislocation characterized by intense cataclastic metamorphism and metasomatism,” which is “characterized by displaced blocks and breccia derived from formations on either side of the abutment” (p. 10). In fact this “zone” is occupied by the reef-margin facies of the reef complexes and is composed of cyanobacterial, stromatoporoid, and coralline limestones that are entirely unmetamorphosed and only rarely faulted (Fig. 42). They also failed to recognize the clear evidence of depositional dips, as proved by geopetal structures in their “homoclines” (i.e., marginal-slope deposits). They claim that “boundaries between all formations and rock units are major sinuous strike faults that dip in the general direction of the homocline” (p. 2). That claim is simply false: the postulated strike faults do not exist.
Logan and Semeniuk proposed that the Devonian carbonate rocks have been subjected to widespread metamorphism: for example, “the development of chlorite, micas and dolomite and the coexistence of dolomite and quartz points to low metamorphic grade” (p. 3) and “we are able to suggest an upper limit of about 300° C to 400° C during metamorphism” (p. 67). That assertion is disproved by many outcrop, petrological, paleontological, and paleo-temperature studies. The rocks have experienced only low-temperature diagenetic alteration (and metasomatism in a few restricted areas). The minor amounts of chlorite and mica that occur in some of these rocks, claimed by Logan and Semeniuk to be evidence of metamorphism up to greenschist facies, are present at only a few places adjoining Precambrian basement. They are detrital, derived by erosion from those Precambrian rocks.
The geothermal history of the area has been analyzed by several authors, using a variety of techniques, as outlined in Playford et al. (2009, p. 124). The evidence indicates that the maximum temperatures reached in exposed and near-surface Devonian carbonates are unlikely to have exceeded 60° to 70° C (Wallace et al. 2002), other than in a few narrow zones (up to about 1.5 m wide) around several volcanic (lamproite) plugs, where temperatures of more than 600° C were reached (Nicoll and Gorter 1984).
Many parts of the reef complexes are stylolitized, but despite claims to the contrary by Logan and Semeniuk, stylolites cannot be mistaken for bedding, even though stylolites often follow bedding. They failed to recognize that strong stylolitization is confined to those parts of the reef complexes that had high porosities when first buried. That porosity was largely destroyed through stylolitization after burial, which yielded carbonate cement that filled void spaces and resulted in compaction of the limestones. Early marine cementation destroyed much of the high primary porosities in reef-margin and reef-flat subfacies, and consequently those deposits were subjected to little postburial stylolitization and compaction. On the other hand, the back-reef subfacies retained its porosity when first buried and, accordingly, it became strongly stylolitized after deeper burial. That gave rise to the characteristic “dished” shapes of many platforms (Playford 1980; Fig. 43).
In espousing their dynamic-metamorphism model, Logan and Semeniuk (1976) have ignored the evidence provided by the well-preserved reef-building fossils in these limestones and the biotic changes that occur in passing from platform through fore-reef into basin facies. The frame-building cyanobacteria, stromatoporoids, and corals in the reef-margin limestones and detrital reef-derived fossils in the fore-reef deposits are clear evidence that these carbonate rocks constitute very well-preserved, unmetamorphosed, reef complexes.
Our knowledge of the Devonian reef complexes has evolved over 130 years, since they were first observed by Hardman in 1884. It was recognized 6 years later that they are Devonian, not Carboniferous, as Hardman had originally thought, but it took many more years before they were recognized as reefs by Wade in 1924.
Much of the early research was conducted in association with oil exploration companies, and that continued after World War II with the work of WAPET. Similar rocks in other parts of the world, notably Canada, have been hosts to large oil fields. However, to date the only Devonian oil discovery has been made at the Blina 1 well, in fractured reefal limestones of the Nullara Limestone (Playford 1982). Other mineral-exploration activities were undertaken for Mississippi Valley Type deposits that had long been known to occur in the reef complexes: epigenetic lead-zinc mineralization was recorded from Narlarla in the Napier Range in 1901 and mined intermittently until 1966 (Ringrose 1989). More extensive epigenetic lead-zinc deposits were discovered in the 1970s and 1980s and have been developed in the Pillara and Emanuel ranges (including the Pillara and Cadjebut mines; Playford et al. 2009). This also resulted in research on the sulfide mineralization and deep-water stromatolite mounds (Playford and Wallace 2001).
Many of the detailed studies discussed above have been carried out by GSWA in collaboration with overseas workers (listed below in the “Acknowledgments”), and one of us (P.E.P.) has also studied Devonian reefal deposits in other parts of the world (Canada, United States, China, Morocco). Hundreds of international researchers have taken part in field excursions to the Canning Basin reef complexes, the first of which was for the International Geological Congress in 1976 (Playford 1976). Details of key excursion localities are given in Playford et al. (2009).
An interesting philosophical outcome of the discovery of deepwater stromatolites on the marginal slopes was to question the “uniformitarian dictum” that stromatolites are solely intertidal in origin, which had resulted from Logan’s (1961) work at Shark Bay. The Devonian discovery led to reappraisal of the modern stromatolites at Hamelin Pool, which showed that those stromatolites are not solely intertidal phenomena but are also widespread in the subtidal environment (Playford and Cockbain 1976, Playford et al. 2013). This is a case in which the past has proved to be a key to the present!
Although a great deal of research on these rocks has now been documented, the potential remains for more research on many aspects of these remarkable reef complexes. Some suggested topics for further research activity include the following:
The Chronostratigraphy Project has shown it is possible to correlate between back-reef, reef, marginal-slope, and basinal deposits using a multifaceted set of constraints. Utilizing the approach and defined stratigraphic framework, correlations can be extended further across the reef complexes and into the subsurface;
Reprocess and reinterpret existing seismic and well data on the Lennard Shelf to improve correlation between the outcropping reef complexes and their equivalents in the subsurface;
Conduct of extensive three-dimensional seismic surveys of the Devonian succession close to outcrop and in deep subsurface areas;
Undertaking detailed studies of known sequence boundaries, as defined by back-stepping events;
Undertake petrographic and field studies of beds that show dissolution of clasts in marginal-slope deposits, especially in various types of breccia;
Undertake detailed studies of the Nullara Limestone, concentrating on cyclicity, patterns of thickness variation, and evidence of sequence boundaries;
Make detailed comparisons with other Devonian reefal carbonates, especially those in Canada;
Undertake more detailed paleontological and geochemical (i.e., elemental) studies across and just below the F/F boundary, within paleomagnetic and isotopic constraints, to further examine the surrounding environmental conditions of the F/F mass extinction;
Undertake detailed structural studies in key areas, especially in the vicinity of the confluence between the Virgin Hills and Sparke Range faults and adjoining the Black Hills Fault;
Undertake detailed studies of the relationships between faulting, platform development, and platform-margin collapse;
Update the systematic paleontology of the invertebrate faunas; and
Undertake detailed studies of renalcids and other microbial reef builders.
It is to be hoped that these magnificently exposed reef complexes will continue to be a focus of geological research, leading to a fuller understanding of their origin, evolution, and depositional environment.
The Geological Survey’s studies have been greatly assisted by geologists from Australia and overseas, who have undertaken studies of the reef complexes and their contained fossils in collaboration with the Survey. Among those collaborators we especially wish to record thanks to Thomas Becker, Scott Brownlaw, Nicole De Kever, Simon Dörling, Ed Druce, Ned Frost, Annette George, Michael House, Neil Hurley, Murray Johnstone, Charles Kerans, Gilbert Klapper, David Lowry, Andrew McManus, Rebecca Mason, Bob Nicoll, Vicki Pedone, Geoffrey Playford, Ted Playton, Doug Smith, Nat Stephens, Dawn Sumner, Linda Tompkins, Phil Tornatora, Malcolm Wallace, Bruce Ward, and Frank Williams.
This account of the reef complexes uses some illustrations from Playford et al. (2009) and Playford et al. (2014), with amendments, where appropriate, and is published with the permission of the Director of the Geological Survey of Western Australia.
Figures & Tables
The Devonian stratigraphic record contains a wealth of information that highlights the response of carbonate platforms to both global-scale and local phenomena that drive carbonate architecture and productivity. Signals embedded particularly in the Middle-Upper Devonian carbonate record related to biotic crises and stressed oceanic conditions, long-term accommodation trends, and peak greenhouse to transitional climatic changes are observed in multiple localities around the world and temporally constrained by biostratigraphy, highlighting distinct and impactful global controls. Devonian datasets also stress the importance of local or regional phenomena, such as bolide impacts, the effects of terrestrial input and paleogeography, syn-depositional tectonics, and high-frequency accommodation drivers, which add complexity to the carbonate stratigraphic record when superimposed on global trends. The unique occurrence of well-studied and pristinely preserved reefal carbonate outcrop and subsurface datasets, ranging across the globe from Australia to Canada, allows for a detailed examination of Devonian carbonate systems from a global perspective and the opportunity to develop well-constrained predictive relationships and conceptual models. Advances in the understanding of the Devonian carbonate system is advantageous considering, not only the classic conventional reservoirs such as the pinnacle reefs of the Alberta Basin, but also emerging conventional reservoirs in Eurasia, and many unconventional plays in North America. The papers in this volume provide updated stratigraphic frameworks for classic Devonian datasets using integrated correlation approaches; new or synthesized frameworks for less studied basins, reservoirs, or areas; and discussions on the complex interplay of extrinsic and intrinsic controls that drive carbonate architectures, productivity, and distribution. The 13 papers in this special publication include outcrop and subsurface studies of Middle to Upper Devonian carbonates of western Canada, the Lennard Shelf of the Canning Basin, Western Australia, and the western USA.