Portal de Periódicos da CAPES

Comment on “From seamount accretion to tectonic erosion: Formation of Osa Mélange and the effects of Cocos Ridge subduction in southern Costa Rica” by P. Vannucchi et al.

2007; Wiley; Volume: 26; Issue: 3 Linguagem: Inglês

10.1029/2006tc002032

1944-9194

David M. Buchs, Peter O. Baumgartner,

Botany and Geology in Latin America and Caribbean

[1] Vannucchi et al. [2006, paragraph 67] report new geological and structural observations of the Osa Peninsula, from which they conclude that the Osa Mélange is a "tectonically disrupted accreted package of oceanic lithologies that are exotic to the overriding Caribbean plate," and "records the accretion of at least two seamount complexes that originated at the Galapagos hot spot between 65 and 25 Ma". They conclude (paragraph 70) also that the Osa Igneous Complex is an accreted seamount system, which probably reached the sea level and suggest that the "Osa Igneous Complex and the Osa Mélange, although parts of the same seamount system, were originally located at different structural levels and followed different deformation paths." [2] Having worked over 20 years in the Osa-Golfito area, we are concerned by the way Vanucchi et al. [2006] describe and interpret the units of the area on the basis of an arbitrary choice of data from the literature and relatively few original observations. We appreciate their structural study which provides new insights into the deformation history of parts of the Osa Mélange. However, with this comment, we would like to clarify some concepts of the regional geology, which we find misleading in the text of Vannucchi et al. [2006]. In addition, we discuss published data and provide new geochemical constraints that are in disagreement with the interpretations made by Vannucchi et al. [2006]. Following the results of Buchs and Stucki [2001], Buchs [2003], Buchs and Baumgartner [2003], we propose that the rocks exposed in NW Osa (the San Pedrillo Unit of Di Marco et al. [1995]) represent the product of (1) erosion and mass-wasting of an arc margin, (2) deposition in a trench, and (3) accretion onto the Osa Igneous Complex. We provide new evidence concerning the arc affinity of some of the igneous blocks embedded within the sedimentary matrix of the Osa Mélange. [3] In order to prevent further confusion concerning the regional geology, we present our geological map of the area, partially reproduced by Vannucchi et al. [2006], which synthesizes all currently available data (Figures 1a and 1b) [Buchs, 2003]. New results of our mapping, paleontologic dating, and geochemical analyses accompanied by detailed interpretations concerning the origin of the Osa area will be presented in a forthcoming paper. [4] Vannucchi et al. [2006] give an account of the geologic framework of the Osa Peninsula and the Golfito area which is mainly based on geochemical analyses and isotopic age dating of exposed igneous rocks [e.g., Hauff et al., 1997, 2000; Sinton et al., 1997; Hoernle et al., 2002]. In doing so, the authors seem to have ignored published and unpublished field surveys that provide detailed petrologic, sedimentologic, and paleontologic data [e.g., Obando, 1986; Di Marco, 1994; Buchs and Baumgartner, 2003]. Preceding the discussion of the origin of the Osa Mélange, we would like to remind the reader of published and unpublished field facts that are in clear contradiction with the picture by Vannucchi et al. [2006]. [5] The various lithologic entities of the Osa-Golfito area have been called differently according to authors. We propose here a simplified nomenclature which reflects the most common usage in recent publications rather than publication precedence. All units figure on the geological map of the area (Figures 1a and 1b). We have constructed this map on the basis of (1) our field work of the last 20 years, (2) a synthetic interpretation of previous publications and unpublished reports (all cited in Figure 1 caption), (3) a detailed, 6 month field survey in the Osa-Golfito area that included mapping 1:5,000 of coastal areas and rivers, and (4) petrologic descriptions of 455 samples and new geochemical analyses of the igneous rocks. [6] The Golfito Complex is referred to as the Golfito Terrane by Di Marco et al. [1995] and Denyer et al. [2006] and as the Golfito ophiolite by Vannucchi et al. [2006]. [7] According to Vannucchi et al. [2006] (see section 9), the inner part of the Osa Peninsula and the area around Golfito (the Golfito Complex and the Osa Igneous Complex) are characterized by an oceanic sequence composed of aphyric pillow lavas, massive basalts, gabbros, plagiogranites, and radiolarian chert. However, our detailed mapping work covering the entire area (Figure 1a) reveals that there are no plagiogranites in the area and that the very minor occurrences of gabbros are restricted to a small part of the Osa Igneous Complex. Radiolarites are restricted to a few scattered outcrops in the Osa Igneous Complex. The Golfito area has been described to be composed of a section of massive and pillowed basalts, interbedded with Campanian-Maastrichtian hemipelagic limestones and overlapped by volcaniclastic deposits that form a several hundred meters thick stratigraphic sequence [Obando, 1986; Di Marco, 1994; Di Marco et al., 1995; Sinton et al., 1997, 1998; Hauff et al., 2000; Buchs et al., 2006]. In view of the absence of ultramafics or layered gabbros, both Osa and Golfito complexes do not meet the criteria to be called "ophiolite" as done by Vannucchi et al. [2006] (compare sections 9 and 10). [8] Similar to the Nicoya Complex (north of Costa Rica), the origin of the Golfito Complex has been related to the Caribbean Large Igneous Province on the basis of the geochemical affinities of the lavas only [Sinton et al., 1997, 1998; Hauff et al., 2000], a view that is accepted by Vannucchi et al. without citing references by Di Marco et al. [1995] that contain opposing stratigraphic and geologic information [Buchs, 2003]. However, strong differences in terms of age, tectonostratigraphy and petrology exist between the two complexes [e.g., Denyer et al., 2006] that call for a different interpretation. Further comments concerning the origin of the Golfito Complex would go beyond the scope of this comment, but we just mention the fact that a synthesis based on both geological and geochemical observations has yet to be carried out. [9] The Osa Igneous Complex is referred to as the Rincon Block by Di Marco et al. [1995] and Denyer et al. [2006]. [10] Vannucchi et al. [2006, section 11] report that the Osa Igneous Complex represents rock bodies formed between 65–60 Ma (Paleocene) and 45–40 (Eocene), citing Berrangé et al. [1989] and Hoernle et al. [2000]. In fact, the radioisotopic dates of the igneous rocks point toward a range of apparent ages that spans 78–48 Ma (Campanian to Eocene), with K/Ar ages of 78–48 Ma [Berrangé et al., 1989] and 39Ar/40Ar ages of 62.1 ± 0.6 and 54.5 ± 1.5 Ma [Hauff et al., 2000; Hoernle et al., 2002]. The paleontological dates based on foraminiferan and radiolarian assemblages indicate similar ages (Campanian to Eocene) [Di Marco, 1994; Di Marco et al., 1995; Diserens, 2002; C. Baumgartner-Mora, personal communication, 2006; A. Bandini, personal communication, 2006]. However, we observed that despite this apparent overall similar age range, igneous and sedimentary ages are locally inconsistent. Important discrepancies exist between the radioisotopic (K/Ar and 39Ar/40Ar) and paleontological ages of sediments interbedded within the lava flows of the Osa Igneous Complex. We argue that at least part of these low-K oceanic tholeiites analyzed in the Osa Igneous Complex suffered from K and/or Ar remobilization during alteration processes and/or seamount accretion. Consequently, at least part of the radioisotopic ages more likely represents hydrothermal and/or tectonic closure ages, rather than crystallization ages of the lavas. In summary, we propose that (1) the age of formation of the bulk of the lavas of the Osa Igneous Complex is limited to the Campanian-Maastrichtian, (2) lesser occurrences of Paleocene (?) to Eocene deposits represent a postvolcanic, preaccretion record of pelagic sedimentation, and (3) middle Eocene pelagic limestones are occasionally associated with small volumes of alkaline pillow lava that pertain to summital portions of seamounts, delaminated and incorporated in the Osa Igneous Complex during accretionary processes [Buchs, 2003]. [11] According to literature, the Osa Igneous Complex is mainly constituted by massive and pillowed basalts (99% of the volume) associated with minor occurrences of cherts, radiolarian cherts, micritic foraminiferan limestones, detrital sediments, rare gabbros, and dolerites [Berrangé and Thorpe, 1988; Di Marco, 1994; Di Marco et al., 1995; Buchs, 2003]. We think that a complete and detailed study of the Osa Igneous Complex has yet to be published. Our new tectonostratigraphic data indicate that the Osa Igneous Complex is a composite unit made of at least four imbricated pieces of seamounts separated by recently reactivated paleodecollement zones and deeper crustal thrusts (Figure 1b) [Buchs, 2003]. No significant amounts of pelagic and hemipelagic sediments are associated with the Osa Igneous Complex [e.g., Berrangé and Thorpe, 1988]. To date, we have not found any shallow water carbonates associated with the Osa Igneous Complex. [12] In contrast to the assertion of Vannucchi et al. [2006, section 11], the Osa Igneous Complex exhibits significant internal deformation and disruption, as demonstrated by frequent overturned position of pillow lavas [Di Marco, 1994] and the exposure of deeper crustal sections. [13] The Osa Mélange is referred to as the Osa-Caño Accretionary Complex by Di Marco et al. [1995] and the Caño-Osa Accretionary Prism by Denyer et al. [2006]. [14] The description of the Osa Mélange by Vannucchi et al. [2006, section 3] reads "the Osa Mélange is a rock assemblage dominated by basalt, chert and limestones" and (paragraph 60) "field indication confirms the little sedimentary component forming the Osa Mélange." These conclusions contradict field observations documented by Di Marco [1994] and Buchs and Stucki [2001] from all coastal and most river outcrops of the outer Osa Peninsula. On the basis of a detailed cartography at a 1:5000 scale and an extensive sampling, it has been demonstrated that sediments form an average of 75% of the volume of the NW Osa Mélange (San Pedrillo Unit) [Buchs and Stucki, 2001; Buchs and Baumgartner, 2003]. Areas exclusively composed of fine-grained (detrital) sediments contrast with areas containing crushed igneous mega-blocks (>10 m in size), which could be confounded with a tectonic mélange, or a disrupted igneous sequence. The high diversity of igneous blocks within small areas has been a strong argument for a gravitational assembly of the mélange [Buchs and Baumgartner, 2003], as well as the sedimentary structures common throughout the San Pedrillo Unit (see section 4.2 for more details). Shallow water calcareous resediments (also reported by Vannucchi et al. [2006, paragraph 20] represent a recurrent lithology of the San Pedrillo Unit, which crops out at Agujitas, Playa Colorada, Punta Campanario, Punta Llorona, and on Caño Island [Buchs and Stucki, 2001]. We also observed large outcrops of these lithologies in the upper reaches of the Rio Tigre (Brazo Derecho [cf. Di Marco, 1994, Figure 42]). Those limestones are systematically embedded within dark tuffites or quartz-bearing sediments and are frequently associated with blocks of dacites or andesites (Figures 2a and 2b) [Buchs and Baumgartner, 2003]. The recurrent co-occurrence of differentiated material and shallow water limestones throughout the San Pedrillo Unit suggests an upper plate, arc-margin origin for the material forming the NW Osa Mélange. This observation is in contradiction with the Vannucchi et al. [2006] ideas of a disrupted oceanic sequence forming the Osa Mélange and an oceanic island provenance for the shallow water limestones. [15] According to Di Marco et al. [1995], the Osa Mélange was divided into the San Pedrillo, Cabo Matapalo, and Salsipuedes units. Vannucchi et al. [2006, paragraph 22] adopt the Di Marco et al.'s concept of the Cabo Matapalo Unit. However, the Salsipuedes Unit, forming the outermost and less deformed unit of the mélange is merged by Vannucchi et al. [2006, paragraph 20] with the San Pedrillo Unit without comment. Di Marco et al. [1995] clearly indicated that this formation cannot be part of the San Pedrillo Unit because the San Pedrillo and Salsipuedes units are separated by the Cabo Matapalo Unit (Figure 1b). They also indicate that the nature of the limestone bodies and the origin of the interbedded fine graywackes which constitute the Salsipuedes Unit require further investigation (i.e., in terms of petrological description and age determination) that is not provided by Vannucchi et al. [2006]. In view of the absence of geochemical constraints on the igneous rocks, the lack of new and refined ages, no petrological description of the sediments, and stratigraphical confusions in the work by Vannucchi et al. [2006], their interpretation concerning the outer Osa Mélange should be considered with precaution. [16] On the basis of the geochemical affinities of the lavas only, it is currently believed that the Osa Igneous Complex represents an accreted aseismic ridge which originated at the Galapagos paleo-hot spot [Hauff et al., 2000; Hoernle et al., 2002]. [17] As mentioned above, no shallow water carbonates have been found in the Osa Igneous Complex. Thus there is no observation supporting the hypothesis that the accreted seamounts in the Osa Igneous Complex may have reached sea level and developed carbonate platforms, as suggested by Vannucchi et al. [2006, paragraph 68]. By now, the study of the stratigraphy of the area has not provided strong constraints concerning the geometrical relationship of the seamounts before the accretion. It is, however, very well illustrated that the edifices suffered dismemberment during accretionary processes. The presence of fossil decollement zones and recently reactivated deeper crustal thrusts within the complex tends to indicate that the pieces of seamounts which constitute the Osa Igneous Complex were not accreted synchronously. The seamounts may have been added to the Central American margin at any time, and possibly over a longer period between the time of deposition of the pelagic sediments on the top of the volcanoes (Campanian to Eocene) and the time of formation of the San Pedrillo Unit (late Eocene), [Buchs, 2003]. [18] The model by Vannucchi et al. [2006, Figure 14] favors the accretion of a "seamount system" (or a "San Pedrillo seamount") ∼20 Ma ago that included the San Pedrillo Unit and the Osa Igneous Complex. However, the ages of formation of the lavas in the Osa Igneous Complex range from the Campanian (∼80 Ma) to the middle Eocene (∼40 Ma) (65–60 Ma to 45–40 Ma according to Vannucchi et al. [2006, paragraph 11]). The difference in age between the youngest and the oldest accreted seamounts of the Osa Igneous Complex is thus of ∼40 Ma (∼25 Ma, using the values of Vannucchi et al. [2006]). If we assume that (1) the velocity of convergence between the Pacific and the Caribbean plates during the Campanian-Tertiary was similar to the present-day velocity (∼90 mm/yr) [Vannucchi et al., 2006, paragraph 61], (2) the seamounts originated at the same hot spot [Vannucchi et al., 2006, paragraphs 11, 61, and 67], and (3) no significant volcanism occurred after the cessation of the volcanic activity close to the hot spot, then the distance separating the seamounts of the Osa Igneous Complex along the hot spot track before the accretion was of ∼3600 km (∼2250 km, using the values of Vannucchi et al. [2006]). This distance appears to be by far too large to allow the accretion of the seamount system over a short period of time during the Miocene (∼20 Ma), as suggested by Vannucchi et al. [2006, Figure 14], and points rather toward (1) a long period of seamount accretion prior to the formation of the San Pedrillo Unit and/or (2) different origins for the accreted seamounts. A detailed observation of the timing of the accretionary processes as proposed by Vannucchi et al. [2006] shows that their model is not self-consistent and is in disagreement with the geological constraints. [19] In summary to this section, we argue that the stratigraphy of the Osa Igneous Complex and the ages of formation of the accreted seamounts are in contradiction with the model proposed by Vannucchi et al. [2006]. [20] There are two fundamental aspects which require to be treated before defining the origin of the NW Osa Mélange: (1) the nature of the fabric and (2) the petrological and geochemical affinities of the matrix and the blocks. [21] Following Raymond [1984], a mélange is a descriptive term for a body of rock, which has a block-in-matrix texture and can be generated by sedimentary processes, tectonic processes, or both. In the case of the Osa Peninsula, most authors propose that the mélange represents gravitational deposits partially dismembered by later tectonic processes [Di Marco et al., 1995; Buchs and Baumgartner, 2003; Vannucchi et al., 2006]. On the other hand, Meschede et al. [1999] suggest that the mélange formed by tectonic processes only and reflects the basal erosion of the Osa Igneous Complex and subsequent underplating of the tectonically eroded material. This interpretation is, however, not supported by geological observations because the compositions of the Osa Igneous Complex and the Osa Mélange are not identical and primary sedimentary structures (such as graded bedding, parallel and ripple laminations) are omnipresent and locally well preserved in the Osa Mélange. [22] The petrological and geochemical affinities of the matrix and the blocks were studied in detail by Buchs and Stucki [2001] at a 1:5000 scale. The San Pedrillo Unit is a highly heterogeneous formation that includes a sedimentary matrix and a wide variety of blocks of sediments and igneous rocks. A complete review of all the lithologies encountered in the mélange is beyond the scope of this comment but they can broadly be classified in three groups: (1) hemipelagic and distal detrital sediments of the matrix, (2) blocks derived from the Osa Igneous Complex, and (3) arc-related blocks and sediments of the matrix. [23] 1. The hemipelagic sediments of the matrix are principally formed by reddish siliceous and/or calcareous mudstones that may contain some basaltic detrital fraction and sometimes abundant radiolarians and/or planktonic foraminifera. These sediments are thought to represent the background sedimentation of a fore arc-trench area. The detrital sediments of the matrix are principally of basaltic composition and contain plagioclases, clinopyroxenes, and opaque minerals [Buchs and Baumgartner, 2003]. [24] 2. The blocks and megablocks of the San Pedrillo Unit that have a seamount origin are identical with respect to lithology, age, and geochemical affinities to the Osa Igneous Complex. All units which constitute the Osa Igneous Complex are represented in the mélange, and thus we infer that the NW Osa Mélange formed after the construction of the Osa Igneous Complex [Buchs, 2003; Buchs and Baumgartner, 2003]. [25] 3. The arc-related sediments are recognized by their contents of detrital quartz, an ashy fraction, and/or the presence of late Eocene resedimented shallow water carbonates. The arc-related igneous blocks vary in size from ∼1 cm to >50 m and are principally represented by dacites that exhibit a typical arc signature (Figure 3). The arc-related fraction represents about 30% of the San Pedrillo Unit. Arc-derived material is intimately associated with hemipelagic sediments, basaltic detrital sediments, and megablocks of the Osa Igneous Complex [Buchs, 2003; Buchs and Baumgartner, 2003]. [26] Temperatures of 200–250°C [Meschede et al., 1999], based on calcite recrystallization and 150–200°C (no method indicated) [Vannucchi et al., 2006], were used to postulate the depth of accretion of the San Pedrillo Unit. However, we argue that the presence of high fluid contents within the subducted material may lead to a nonrepresentative estimation of the temperatures because pervasive fluids at different levels within the wedge may have had a significant effect on the geothermal gradient and calcite recrystallization. In absence of any detailed study of the low-grade metamorphism or hydrothermal processes, it seems problematical to determine the exact depth of accretion of the Osa Mélange on the basis of thermal gradients. [27] The sedimentary facies of the San Pedrillo Unit indicate that large amounts of detrital material deposited rapidly in the trench during catastrophic events [Buchs and Baumgartner, 2003]. Those sediments dramatically increased the sediment pile in the trench making it difficult to subduct them to a great depth [e.g., Le Pichon et al., 1993]. The most probable case is that most of the sediments were accreted at a shallow depth, directly onto the Osa Igneous Complex, even if the downgoing plate had topographic heterogeneities to accommodate part of the sedimentary supply. [28] In summary to the preceding sections, we observe for the NW Osa Mélange (San Pedrillo Unit) that (1) the fabric is consistent with an initial formation primarily controlled by sedimentary/gravitational processes, (2) the sedimentary deposits were subsequently tectonically dismembered, presumably during accretion and later effects of the Cocos Ridge subduction, in such a manner that the original stratigraphy is preserved at a large scale, (3) the igneous blocks have very distinctive geochemical affinities and are entirely embedded within the sedimentary matrix of the mélange, and (4) the affinities of the detrital fraction (including megablocks) show that the mélange formation is a consequence of the erosion of (1) all the units represented in the Osa Igneous Complex and (2) an emerged volcanic arc and carbonate platforms lying on the edge of the arc. These observations do not support the model of Vannucchi et al. [2006] for the main reason that it is in disagreement with the pervasive sedimentary association of arc-related and seamount-related material within the NW Osa Mélange. [29] The presence of material originating from the Osa Igneous Complex and a volcanic arc within the San Pedrillo Unit indicate that the NW Osa Mélange formed close to an arc system, most probably after the accretion of the Osa Igneous Complex [Buchs, 2003]. As a consequence, it is obvious that the Osa Mélange is not exotic to the overriding Caribbean plate and is not part of the "accreted seamount system" of Vannucchi et al. [2006]. At present, there are no geological or geochemical constraints indicating that the units forming the Osa Igneous Complex are part of the same group of seamounts, nor were accreted synchronously [Buchs, 2003]. [30] In our model [Buchs, 2003; Buchs and Baumgartner, 2003], we propose that the NW Osa Mélange formed by the accretion of a an accumulation of detrital material in an arc-trench system that was gravitationally derived from the earlier accreted Osa Igneous Complex and an arc margin (Figures 4a–4f). This is in opposition to an accretion of purely seamount-derived material directly from the subducting plate to the overriding plate [Vannucchi et al., 2006] (Figures 4g and 4h). On the basis of available and new paleontological dates of the Osa area, we determined that the bulk of the detrital material was generated during the late Eocene and presumably rapidly accreted thereafter. Hypotheses concerning the causes of the margin collapse will be discussed in forthcoming paper. A synthesis of the differences between our model and the model of Vannucchi et al. [2006] is illustrated in Figure 4. [31] In the light of the geological and geochemical constraints presented here, we wonder how the detailed structural observations by Vannucchi et al. [2006] may help to characterize the accretionary processes and depth of subduction of the NW Osa Mélange. In a constructive way, we address this question to Vannucchi et al. [2006]. [32] We acknowledge the quick review by anonymous reviewers. This study has highly profited from micropaleontological dating by A. N. Bandini, C. Baumgartner-Mora, M.-O. Diserens, S.-J. Jackett. Geochemical analyses were performed at the Centre d'Analyse Minérale (University of Lausanne) by J.-C. Lavanchy and at the Institute of Mineralogy (University of Lausanne) with the help of F. Bussy and A. Ulianov. Thin sections were made by L. Nicod at the Institute of Geology (University of Lausanne). We are thankful to R. Arculus and J. Hernandez for their constructive criticism during field work and geochemical interpretations. We greatly appreciated the hospitality and enthusiasm at Marenco Beach and Rain forest Lodge. This study was carried out in the framework of a research project of the Swiss National Science Foundation (00021-105845). Earlier field studies were supported by the Herbette Foundation (University of Lausanne).