The Pacific margin of Costa Rica, offshore of the Nicoya peninsula (Fig. F1), is bathymetrically smooth. The margin's surface is covered by a sedimentary apron (Fig. F2) that varies in thickness from 0.5 to >2 km (Shipley et al., 1992). The apron overlies a deformed prism, composed of relatively high velocity material (Christeson et al., 1999; Ye et al., 1996). Both the prism and the apron are cut by landward-dipping thrust faults in the middle and lower slope region (Fig. F2), which are termed out-of-sequence thrusts (Shipley et al., 1992). Researchers on Alvin dives on this margin observed that active fluid venting occurs where these thrusts cut through to the surface (McAdoo et al., 1996; McIntosh and Silver, 1996). The lowermost 5 km of the margin is a deformed sedimentary wedge, also with landward-dipping thrusts. At Ocean Drilling Program drilling Sites 1040 and 1043, we documented that the material composing the wedge is not derived predominantly from the incoming sediment on the Cocos plate. Rather, it is more consistent with material recovered at Site 1041 that penetrated most of the sedimentary apron about 8 km upslope from Site 1040 (Kimura, Silver, Blum, et al., 1997). Postcruise studies by Valentine et al. (1997) have shown that the wedge material is very low in 10Be, indicating a relatively old age for the deformed wedge.
The age-depth curves from reference Sites 1039 and 1040, 1.6 km upslope from the toe, show the same distribution, with zero age sediments directly underlying the décollement. Careful examination of resistivity and gamma ray logs from logging while drilling (LWD) shows that, wiggle for wiggle, all of the material at Site 1039 is present (somewhat compressed) at Site 1043, 0.5 km upslope from the toe, with the exception of the uppermost 9 m of sediment (Saito, 1997). Even this amount cannot be considered as scraped off, because Site 1039 has 5 m of terrigenous turbidites capping the section. A 3.5-kHz seismic line between the two sites shows that these sediments are removed from the trench axis at the toe, likely due to current scour (Kimura, Silver, Blum, et al., 1997). Moritz et al. (in press) applied the method of a genetically trained neural network to the comparison of LWD data at Sites 1039 and 1043. Excluding the top 5 m of turbidites at Site 1039, as discussed above, they found that no sediment was missing between the reference site and the underthrust sediments at Site 1043.
One hole was drilled through the apron into the prism material at Site 1042 (Fig. F1). At this site we recovered at least 60 m of prism material beneath the apron. Microfossils recovered from the prism showed several age inversions, indicating the presence of several thrust sheets, with rock ages ranging from early to late Miocene age. These thrust sheets contained calcarenitic breccias and fragments of mafic rocks and chert (Kimura, Silver, Blum et al., 1997), indicating a provenance different from that of the incoming Cocos plate stratigraphy. These rocks are consistent with exposures onshore, with some derivation from the Nicoya complex and related rocks. In addition, measurements of seismic velocities within these breccias gave values of 4 km/s (Kimura, Silver, Blum, et al., 1997), consistent with the results of wide-angle refraction studies (Christeson et al., 1999). The drilling results are also consistent with the seismic reflection data, showing a significant amount of thrusting in the prism beneath the lower slope and especially a concentration of faulting at the intersection of the prism and the deformed sedimentary wedge (Fig. F2) (see also McIntosh and Sen, 2000).
When we combine the drilling results with the seismic reflection and refraction data, a view emerges of the prism material as follows. The upper slope part of the prism has a high seismic velocity, with no discernible change between the slope region and the onshore area underlain by the Nicoya complex. Thus the upper part of the prism very likely comprises Nicoya complex rocks. Deformation in the upper slope region is by normal faulting (McIntosh et al., 1993), whereas the middle and lower slope regions are characterized by thrust faulting. Seismic velocities decrease downslope, implying that the thrust faults either break up the prism and diminish elastic properties through fracture and alteration or that lower velocity sedimentary material is interthrust with the higher velocity rocks. In the event of the latter alternative, the low-velocity material could consist of sediment that was either deposited originally on the slope apron or underplated beneath the prism, then thrust into it. In addition, the nature of the prism material may change downslope, with more of a detrital (breccia) component deposited seaward of the Nicoya igneous rocks. It is likely that a combination of the above mechanisms acts to vary the measured seismic velocities.
The underthrust Cocos plate stratigraphy plays a significant role in the evolution of the margin. Although offscraping does not come into play at the toe of the slope, McIntosh and Sen (2000) have discovered evidence (also reported by Shipley et al., 1992) for some underplating at depth in the seismic records. One example is subtly illustrated in Figure F2, just to the right of the arrow pointing to "underthrust sediment." Here a mass of material several hundred meters thick and about 1 km wide appears to be underplated beneath the deformed wedge. The underthrust sediment, especially the upper hemipelagic layer, is significantly reduced in thickness from that in the trench because of the load of the deformed wedge and the fact that the uppermost high-porosity strata are all underthrust beneath the toe (Kimura, Silver, Blum, et al., 1997). McIntosh and Sen (2000) have carefully mapped the variation in thickness of the hemipelagic layer with depth beneath the wedge on depth-migrated seismic sections. They find significant thickening of this layer several kilometers landward of the trench axis, an observation explicable by internal deformation of the underthrust hemipelagic sediment. Strongly supporting this observation is the finding during Leg 170 that steeply dipping strata were measured in the underthust hemipelagic layer at both Sites 1040 and 1043 (Kimura, Silver, Blum, et al., 1997). The zone of anomalously steep dips includes the entire hemipelagic section at Site 1040 and the upper part of the section at Site 1043. There is some indication of landward dips cutting the strata in the underthrust section in Figure F2.