Reconstruction of Topography and Related Depositional Systems during Active Thrusting

Reftable reconstruction of former topography in deformed regions is commonly difficult, due to degradation of former erosional and depositional surfaces. In contrast to most modem landscapes, however, ancient localities can sometimes provide clearer insights on subsurface geometries of deposition, deformation, and erosion and on theft variations through time. In some exceptional circumstances, ancient depositional sequences are preserved in direct juxtaposition with the structures that controlled their geometrical and sedimentological character. We describe here the evolving topography and depositional responses caused by the late Eocene growth of a detachment fold and related thrusts in the southern Pyrenees. Topography within these deforming systems can be reconstructed on the basis of (1) relief associated with paleovalleys, (2) geometric relationships of syntectonic strata with adjacent structures, and (3) relief of hanging walls above depositional or erosional surfaces of the same age. Onlapping, offiapping, and overlapping stratigraphic relationships are interpreted in the context of the relative rate of sediment accumulation versus the rate of uplift of the crest of the fold. In the study area, two contrasting fluvial systems provided sediment to the deforming area: a large longitudinal system, flowing parallel to the fold axes and carrying detritus from the distant hinterland, and a smaller transverse system that carried locally derived clasts. During fold growth, syntectonic sedimentary beds (growth strata) were progressively rotated in the forelimb of the fold. Proximal unconformities developed in the forelimb growth strata, when accumulation rates were low. Topographic relief on the backlimb of the growing fold caused transverse paleovalleys (> 150 rn deep) to be incised at high angles to the fold axis. A switch from incision to infilling of the paleovalleys appears to be controlled by relative rates of subsidence, sediment supply and accumulation, and uplift. During an interval of rapid accumulation and low rates of subsidence and uplift, the effects of rising local base levels propagated up the transverse valleys, where they initiated backfilling of the paleovalleys. As deformation began on an adjacent, more hinterlandward thrust, waning growth of the detachment fold permitted depositional overlap of its crest, as sedimentation shifted toward the hinterland. Subsequently, as the new footwall was folded, longitudinal rivers filled the space formerly occupied by transverse rivers, and a new detachment fold grew in the very shallow (<25 m) subsurface. Although similar examples are scarce in the geological record, the synthesis from this Pyrenean locale illustrates how stratal geometries, reconstructed fiver patterns, precise stratigraphic …


Introduction
can be defined by the height differences between the depositional surfaces and preserved erosional topography nearby.

Geologic Setting
The South Central Unit (SCU) [Sdguret, 1972] comprises an imbricate system of emergent thrusts (from north to south: the Boixols, Montsec, and Serres Marginals thrusts, Figure 1) which display an overall piggyback sequence of thrusting during latest Cretaceous to early Eocene times.Subsequent southward transport over the foreland deposits is recorded by middle Eocene to Oligocene syntectonic strata [Vergds and Burbank, 1994].Coeval with this southward transport, older hinterland thrusts were reactivated, in both the central [Martinez-Pe•a and Pocov [, 1988;Mu•oz, 1992;Puigdefdbregas et al., 1992] and eastern parts of the SCU (Figure 1) [Burbank et al., 1992;Verggs and Muaoz, 1990].
The study area is located along the eastern boundary of the SCU, where a set of oblique thrusts merge northward into a single thrust: the Segre thrust.The Oliana anticline is a 15-kmlong, doubly plunging structure located in the footwall of and oblique to the Segre thrust.Formed by stacking two slices of Paleogene strata, the anticline is an allochthonous structure that was transported 10 km to the south across a decollement localized within the upper Eocene Cardona salt [Vergds and Mu•oz, 1990;Vergds et al., 1991].To the northwest of the Oliana anticline and coeval with its growth, synorogenic conglomerates that overlap and are cut by thrusts demonstrate that the Mesozoic rocks of the SCU in the hanging wall of the Segre thrust are deformed by break-back thrusts (i.e., a sequence of thrusting that becomes younger toward the hinterland) and related folds.Details of the thrust geometry, sequencing, age, and rates of shortening are given by Vergds and Muhoz [1990] and Burbank et al. [ 1992].The present study describes the landscape evolution associated with growth of a detachment fold within the Setres Marginals thrust sheets and later shortening on the Montsec thrust.Both the fold and the hanging walls of these thrusts include Mesozoic and Paleogene strata (Figures 3 and 4).

Methodology
The study area was mapped in detail on aerial photographs and topographic maps at scales of -1:16,000 and 1:25,000, respec- fively.Sections were measured and described through each of the major depositional units, and data on facies, paleocurrents, and provenance were collected.The present geometry of former erosional and depositional surfaces was recorded and later restored to a predeformational position with respect to bedding surfaces that were assumed to have been essentially horizontal at the time of deposition.Samples for paleomagnetic studies were collected where fine grained strata were sufficiently abundant and sections were sufficiently long.These samples were processed through multiple levels of thermal demagnetization: a procedure used successfully in previous studies to create magnetostratigraphies in correlative strata on the flanks of the Oliana anticline (Figure 3).(For details of the magnetostratigraphic procedures, see Burbank et al. [1992].)Interpretation of the growth strata is based in part on models that predict stratal geometries as a function of the relative rates of sediment accumulation versus crestal uplift (Figure 5).Crestal uplift is defined as the structural uplift plus or minus any deposition or erosion, respectively, over the crestal area of a fold.Crestal uplift thus defines changes in position of the highest surface of the fold and is differentiated from bedrock uplift [England and Molnar, 1990].The reference level for measuring crestal uplift is either the position of correlative folded strata in the adjacent syncline or the base of the syntectonic strata in the adjacent basin.Although here we use crestal uplift rates, one could also interpret onlap-offlap geometries in the context of uplift rates of a site on the flanks of a fold.

Syntectonic Deposition
The upper Eocene to lower Oligocene stratigraphic succession [Burbank et al., 1992] is dominated by four, primarily conglomeratic sequences that are interleaved with various thrusts and deformed by the Oliana anticline (Figures 3 and 4 (2) the position of correlative marker beds found in both the anticline and the adjacent syncline.When rates of accumulation are consistently greater than the rate of crestal uplift, overlap will occur (Figure 5b), whereas lower rates of accumulation versus uplift lead to offiap (Figure 5c).Reversals in the relative magnitude of these rates causes a switch in the bedding geometry (Figure 5d).Onlap (Figure 5e) occurs following offiap and a change to more rapid accumulation rates.
occurrences lie between the hanging wall of the Montsec thrust and the backlimb of the Oliana anticline which was also growing during deposition (Figures 1 and 4).Growth syncline.Unit 2 is at least 250 m thick within the axis of the growth syncline (Figure 7 Age: The results from 16 class I and II magnetic sites following stepwise thermal demagnetization indicate that at least three and probably six magnetic reversals are encompassed by unit 2 (Figure 8).In the context of the previously determined magnetostratigraphic framework for units 2 and 3 on the SE flank of the Oliana anticline [Burbank et al., 1992], this reversal pattern can be interpreted (Figure 8) to indicate that unit 2 was deposited between ~36.7 and 35.0 Ma [Cande and Kent, 1992 5), although the crestal uplift rate of the anticline was likely to have been considerably higher.
To the NW, unit 2 is juxtaposed against and incorporated in the forelimb of the detachment fold (Figure 6).Clear onlapping of the conglomerates with the pre-Eocene strata is seen along much of the exposed contact.During growth of this anticlinal fold, formerly subhorizontal depositional surfaces within unit 2 were progressively rotated, such that some older surfaces are presently overturned in the proximal parts of the growth syncline.A clear hinge line separates (1) the central area of unit 2 deposition in which stratal thicknesses are generafly constant and beds are tabular from (2) the region adjacent to the anticfine in which stratal thicknesses abruptly taper toward the anticline.Because these tapered strata are lateral continuations of adjacent tabular strata, comprise well rounded clasts, are largely fluvial in origin, and contain basement clasts even in many sites close to the fold, they represent part of an axial fluvial system flowing parallel to the fold axis and can be considered to have had subhorizontal upper surfaces at the time of deposition (with respect to the NW-SE cross section), rather than being sloping transverse fan surfaces.
Reconstructed history of deformation and deposition: Our interpretation of the magnetic data implies that sediment-accumulation rates varied strongly during deposition of unit 2. In comparison to relative uplift rates on the flank of the fold, high rates of aggradation (~25 cm/kyr) during the initial phases of deposition promoted onlap of the detachment forelimb.The progressively rotated bedding surfaces of the upper 65% of unit 2 and the geometry of their contacts define the growth of the detachment fold and related erosion and deposition.They demonstrate that during a 1-m.y.-long interval, the fold's forelimb progressively steepened and overturned in response to continued shortening.As the anticfine grew, its crestal area and those pregrowth strata on its backlimb (i.e., all strata that were not manfled by syntectonic sediments) were subjected to erosion.•......:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:  The topmost part of the valley fill below the base of unit 3 is a 20-to 25-m-thick succession of thin (<40 cm) sandstones (20% of total thickness) and yellow-orange and pink siltstones.These strata are extensively mottled and burrowed, and they contain abundant rootlets and pedogenically altered sandstones.This fine-grained sequence drapes across the top of both palcovalleys and may also correspond with a zone lacking outcrops at the unit 2/3 boundary in the growth syncline.
The sedimentological data indicate an alluvial fan of local origin filled the medial part of the major palcovalley.Persistent imbrications, low-matrix content, large lateral extent of beds, and limited stratal thicknesses suggest that deposition occurred from unconfined fluvial flows, some of which are likely to have been hyperconcentrated [Smith, 1988].The boulder clusters are interpreted as bar cores around which low-angle accretion occurred as the bar expanded [Hein, 1984].
Along the western valley margin, the fine-grained nature of the strata, abundant bioturbation and soil development, and occasional thin limestones suggest relatively slow deposition on a low-gradient floodplain.Rapidly thinning conglomeratic units, 1-3 m thick, encroached into this marginal environment infrequently.
Given the close proximity of these two sections (-400 m), the abrupt facies changes between them indicate that the lateral margin of an alluvial fan is preserved here (Figures 9 and 10).Highly energetic depositional environments are present in the valley center, whereas ponded and tranquil deposition characterizes the region between the fan and the valley wall.
The extensive palcosol-dominated succession that caps unit 2 suggests that an interval of quiescence prevailed prior to the initiation of deposition of unit 3. The absence of conglomerates further suggests that the drainage system was being reorganized, possibly in response to the initial transfer of thrust deformation to the more hindward Montsec thrust.
Geometrical relationships: Near the western margin of the major palcovalley, the basal strata of unit 2 have been rotated during deposition.Over a stratigraphic span of 20 m, bedding dips decrease from vertical to -55 ø to the NW.These beds were deposited coevally with some of the syntectonic strata in the forelimb of the growing detachment fold.The rotated beds can be interpreted as a wedge-shaped "progressive unconformity" [Riba, 1976] that opens toward the hinterland.This back-tilting is interpreted to result from flexural slip folding during tightening of the syncline in the pre-Eocene strata which underlie the margin of the major palcovalley.Uniform dips of beds above the rotated strata suggest that differential folding slowed or ended midway through the filling of the palcovalley.
Given that the palcovalley represents at least 150 m of incision near the trace of the Montsec thrust, it is not surprising that considerable erosion also occurred across the crest of the growing detachment fold where traversed by the palcovalley.During this incision, uplift of the core of the detachment fold was apparently too rapid to permit permanent accumulation of syntectonic strata within the palcovalley.
Because individual strata and groupings of strata can be physically traced from the paleovalley's western margin toward its center, it is possible to compare lateral changes in stratigraphic thicknesses and apparent sediment-accumulation rates.Four reference horizons serve to delineate these rate changes: the base of the palcovalley; the base of "unit A"; the top of "unit F", and the conglomerate (the "passing conglomerate"; Figures 9B and 10) that crosses the top of the minor palcovalley and can be traced toward the center of the major palcovalley.Comparison of these markers shows that there was a clear-cut shift of the zone of rapid accumulation from the valley center toward its margin.Below unit A, a much greater thickness of strata accumulated in the central part of the palcovalley.Part of this is simply due to passive infilling of an eroded landscape.Traceable beds which thin toward the valley margin, however, indicate that there was also greater subsidence in the valley center.This trend continues up through unit F, during which time there is 2-3 times more accumulation in the valley center than near its western margin (Figures 10 and 11).This interval coincides with the previously described progressive rotation of the basal strata along the valley wall, such that the contrasts in accumulation probably result from differential uplift of the valley wall during folding of the Serres Marginales hanging wall.
Neither the observed 7 ø taper that results from thinning of groups of beds nor the rapid change from a tabular bed geometry in the valley center to a strongly tapering one toward the margin are consistent with the expected bedding geometry and topographic shape of an unperturbed alluvial fan [Wallace, 1978;West, 1991].It therefore appears that folding at this scale can have dramatic effects on sediment accumulation across short (~400 m) distances and that differential subsidence confined the fan to the valley center during the early part of the valley filling.
Between unit F and the "passing conglomerate", the opposite accumulation trend prevails: there is more accumulation toward the valley margin than in its center (Figure 10).In the lower part of this sequence, there appear to be nearly equal accumulation rates, but in the later stages during predominantly finer grained deposition, the valley-margin rates appear to be 2-4 times greater than in the center.This suggests that, during the later stages of unit 2 deposition, deformation was shifting to a more hindward location, i.e., the Montsec thrust [Burbank et al., 1992], and beginning to deform the proximal part of the basin adjacent to the One remarkable aspect of unit 3 is the incorporation of a megaclast of Cretaceous limestone (Figure 9).This very large block (-350 m long and > 100 m thick) apparently slid from the hanging wall to the footwall of the Montsec thrust.The base of this megaclast is not exposed, but abutting its margins are cones of unsorted, angular clasts of locally derived monolithologic breccias that are incorporated into ill-defined beds dipping steeply away from the block.At higher stratigraphic levels, the megaclast is onlapped and overtopped by tabular units of conglomerates and siltstones that display no significant thickness changes as they approach the depositional contact with the block.The palcovalleys represent the only significant transverse elements in the evolving landscape that are still preserved.These valleys are oriented perpendicular to the structural trend and served as conduits carrying locally derived detritus to the longitudinally oriented, larger river (section B-B", Figure 2).The volumetrically small contribution (<10%; Figures 2, 3, and 4 The incision of the larger transverse palcovalley through the crest of the growing Serres Marginals anticline caused a locally important unconformity to be generated within unit 2 in the growth syncline.This resulted in recognizable along-strike variations in the geometric relationship between syntectonic strata and the folded pre-Eocene strata.The deeper incision that occurred along the major palcovalley caused juxtaposition of unit 2 with eroded limestone in the backlimb of the fold (section B-B', Figure 7), rather than overlapping the forelimb, as is seen farther to the west (section A-A', Figure 7).
The switch fxom incision to deposition along the palcovalleys occurred while the detachment fold was growing and the strata of unit 2 were filling the growth syncline.We assume that because the transverse valleys were tributaries to longitudinal river in the growth syncline (Figures 2 and 12a), the position of the top of the fill in the syncline represented the local base level for the transverse paleovalleys.If the top of the fill is raised above the level of the incised bedrock of the paleovalley, sediments will aggrade within the paleovalley.If, on the other hand, the depositional surface in the growth syncline is low with respect to the bedrock of the paleovalley and if the transverse sediment supply is not high, then continued incision of the transverse valley will occur.Deposition or erosion in the paleovalleys can therefore be interpreted in the context of changes in (1) the balance between the rate of sediment supply to the growth syncline from longitudinal rivers and the rate of subsidence of the syncline and (2) the rate of crestal uplift of the detachment fold versus sediment supply and stream power in the transverse paleovalleys.In the context of these conceptual interactions, incision in the transverse valley is promoted by high rates of subsidence in the growth syncline and of crestal uplift in the fold, low rates of sediment supply to the transverse valley and of accumulation in the growth syncline, and high stream power in the transverse valley.In the study area, the deepest incision of the paleovalley and the development of an unconformity in the growth syncline near the mouth of the transverse valley appear to have occurred when sediment-accumulation rates were at their lowest levels (Figures 2, 7, and 8).Aggradation and backfilling in the paleovalleys are interpreted to be promoted by rapid accumulation in the growth syncline, high rates of sediment supply to the transverse valley, decreases in subsidence in the growth syncline and/or of uplift in the crestal region of the detachment fold, and decreases in stream power in the transverse valley.In the context of the magnetic chronology (Figure 8), the doubling of accumulation rates after 35.7 Ma and a relative slowing of crestal uplift rates in the detachment fold appear to correspond with the switch from incision to infilling in the larger paleovalley.
The minimum topographic relief that can be documented on the larger paleovalley is -150 m, which represents the stratigraphic separation of the base and the top of the paleovalley fill.The average slope of the preserved valley wall is about-20 ø .During the ~l-m.y.-long interval when the paleovalley was filling, erosion of the interfluves would have reduced the original relief on the paleovalley to that observable today.
In the case of unit 2, nearly all of the hanging wall topography developed during growth of the detachment fold and during tightening of the previously formed syncline which was inherited from earlier deformation in the Eocene (Figure 2 Comparison of the reconstructed geometries of the Oliana region with those associated with presently active structures serves to emphasize the importance of relative rates of tectonic, erosional, and depositional processes in controlling the evolution of the landscape.In a structural position analogous to the Oliana anticline with respect to the Ebro basin during the latest Eocene, Wheeler Ridge in the southern San Joaquin valley of California represents the outermost major fold [Medwedeff, 1992] in an active fold-and-thrust belt that borders a large depositional basin.Wheeler Ridge is an eastward plunging anticline that dives below the modern depositional plain at its eastern end (Figure 12b).Its hanging wall has been cut by transverse streams that are localized by transverse faults [Medwedeff, 1992].Where rates of hanging wall uplift have exceeded rates of incision of the anticline, former water gaps have been converted to wind gaps.On the distal, northern flank of Wheeler Ridge, short-radius (generally <1 km), locally sourced alluvial fans are being actively deformed in their proximal parts by the growing fold.To the south of the fold crest, the depositional plain between the fold and the Pleito thrust is progressively incorporated into the rapidly uplifting backlimb of the growing anticline, where it is attacked by erosional processes.The emergence of Wheeler Ridge above the depositional plain has strongly perturbed the previous depositional systems.Small, transverse streams that had fed into the site of the present wind gap have been diverted to the east, where they join others and are either focused into the water gap or are diverted around the plunging nose of the anticline.The topography on the large alluvial fans near to Wheeler Ridge suggests that they also have been deflected around the growing anticline.On the northem side of the anticline, which is downstream with respect to the regional drainage system, two subregional alluvial fan systems converge (Figure 12b) toward a zone of reduced accumulation: a "depositional shadow zone."To the south of Wheeler Ridge, the active Pleito thrust uplifts the more proximal parts of the thrust belt and causes transverse streams to incise into its hanging wall.
At two scales, Wheeler Ridge displays geometries that are analogous with those reconstructed at Oliana.At the small scale, the transverse valleys in Wheeler Ridge are comparable to the Oliana paleovalleys during stages in which local uplift rates of the fold crest surpassed the accumulation rates on the footwall.
At the large scale, Wheeler Ridge presents a major topographic barrier that deflects major drainage systems around it.Similarly, the Oliana anticline acted as a divisive barrier to the ancestral Segre River (Figure 12a) which was diverted either into the broad foreland to its southeast which lacked many topographic barriers or into the structural moat along its northwest flank (analogous to the separation between the Pleito Thrust and Wheeler Ridge).
There are several important differences between Wheeler Ridge and Oliana.First, the major drainage is not open to the west, because the uplifted Pleito hanging wall and Wheeler Ridge intersect toward the west and their mutual uplift rates exceed the rate of aggradation.Consequently, Wheeler Ridge cannot divert a major fiver system to either flank, as happens at Oliana.Second, whereas the depositional area above the submerged eastern fold nose probably provides an excellent model for overlapping syntectonic geometries (Figure 5b), for much of the length of Wheeler Ridge, the rate of crestal uplift (3.2 m/kyr) is considerably greater than the rate of sediment accumulation (1.8 m/kyr) adjacent to the anticline [Medwedeff, 1992].As a result, rather than developing gradually tapering syntectonic beds that are incorporated and recognizable in the forelimb as is seen in the detachment fold at Oliana, syntectonic strata at Wheeler Ridge pinch abruptly against the forelimb and display such strong offlap (Figure 5c) that they can not clearly be seen as incorporated into the growing fold.Third, the presence of readily eroded marl in the core of the Oliana anticline inhibits recognition, on the distal side of the fold, of fans with locally derived clasts that would be analogous to those at Wheeler Ridge.
When both the differences and similarities between Oliana and Wheeler Ridge are considered, they define a spectrum of responses to active contractional deformation.At Wheeler Ridge, an ergodic substitution of space for time serves to delineate likely responses of the landscape to varying rates of uplift and aggradation.These spatially variable rates produce depositional and topographic geometries that mimic many of those that result from the temporally varying rates and processes, such as those defined at Oliana.The modern environment of Wheeler Ridge serves to define more clearly surficial relationships between structure, topography, and sedimentation, whereas the ancient setting of Oliana permits one to view the depositional, erosional, and structural geometries that are presently obscured beneath the Earth's surface at Wheeler Ridge. 5. When aggradation is rapid in comparison to uplift, growing structures and any preexisting topography can be onlapped or overlapped by syntectonic strata.At Oliana, accumulation rates of -25 cm/kyr are sufficiently rapid to create strong onlap.Near equality in rates of accumulation and uplift lead to conspicuous thinning and folding of syntectonic strata into a growth syncline (progres siv e unco nformity).

Conclusions
6. Large-scale geometries of strata within growth synclines also depend on the erosion of the crestal area of the adjacent fold.When the crest remains intact, continuous deposition and folding are often recorded by the syntectonic strata in the forelimb.In contrast, when the crest is breached, secondary unconformities can develop within the growth synclines due to lowered transverse gradients along paleovalleys, especially during times of low accumulation and rapid subsidence in the syncline.
7. Very similar structural geometries can be found through a comparison of the Oliana area with the presently deforming Wheeler Ridge in southern California.This modern analog of a plunging anticline impinged upon by hindward thrust sheets displays many of the same features that have been reconstructed for the Oliana area, including structurally controlled transverse paleovalleys and structurally diverted drainages.The main differences between these two settings result from the more rapid rates of crestal uplift at Wheeler Ridge which results in pronounced offlap of many syntectonic strata and in the development of pronounced topographic barriers to flow.Whereas modern analogs can display unambiguous modern geomorphic relationships, the record of their previous deformational and depositional history is often obscured in the subsurface.Although geomorphic processes and topography at ancient sites can only be reconstructed, rather than observed, stratigraphic and structural exposures provide the opportunity to document the temporal evolution of the structure and the landscapes related to its deformation.

Figure 1 .Figure 2 .
Figure 1.Simplified geologic map showing major thrusts and stratigraphic units within the Oliana area.The locations of the master magnetostratigraphic section [from Burbank et at., 1992] and the studied area (F?gure 3) are shown.Inset shows a generalized map of the Pyrenees.The study area is located astride the eastern oblique margin of the South Central Unit (SCU).

Figure 3 .Figure 4 .
Figure 3. Geologic map of the syntectonic strata (units 1-4) associated with the Montsec and Serres Marginals thrust sheets (shaded).Growth synclines, as depicted by strike line mapping, occur in the footwall of the Montsec thrust and in the forelimb of the detachment fold in the Setres Marginals hanging wall.The bases of the paleovalleys and the breached crest of the detachment fold are shown.The large arrow depicts the trend of the now-eroded palcovalley across the detachment fold.Locations of the magnetic section and various cross sections are shown.
Figure 5. (a) Nomenclature for a detachment fold and growth strata deposited during folding.(b)-(e) Models of overlap, offiap, and onlap.Predictable geometries of synteetonie strata result from contrasts in the relative rates of erestal uplift versus coeval rates of accumulation.Crestal uplift is measured with respect to either (1) the base of the synteetonie strata adjacent to the fold or clasts are well rounded and are heterolithic.Although dominated by clasts of Mesozoic carbonates that could be locally derived, unit 2 within the growth syncline also contains basement clasts derived from the Pyrenean "Axial Zone" >30 km to the north.The conglomerates appear to be of wholly local origin only in localities within a few meters of the growing detachment fold.Palcocurrent directions (N = 113) based on clast imbrications, cross bedding, and numerous elongate "gutters" on the base of many conglomerate units (Figure 8) indicate that flow, across much of its extent, was to the SW, i.e., a longitudinal or axial flow direction [Burbank and Beck, 1991; Eisbacher et al., 1974] parallel to the traces of the local thrusts and fold axes.At the base of the succession on the folded forelimb, poorly bedded, unsorted, monolithologic breccias of angular clasts of Garumnian limestones in jig-saw-like geometries are present in places.In other sites, crudely bedded, subangular to subrounded conglomerates with large clasts of Garumnian limestones as much as 1 m in diameter occur.

?Figure 6 .
Figure 6.Overview of the upper i00 m of unit 2 in the growth syncline and the overlying conglomerates of units 3 and 4. Progressive tapering and steepening of each stratigraphic packet in unit 2 delineates the growth syncline adjacent to the folded forelimb of Garumnian limestones (visible on the right skyline).

Figure 7 .
Figure 7. Simplified cross sections of unit 2 in the growth syncline.The location of the magnetic section is projected into the plane of sections.Following early onlap of syntectonic conglomerates on preexisting and slowly deforming footwall topography, continuous folding of the unbreached anticline led to an overturned, wedge-shaped progressive unconformity (A-A').Erosional breaching of the detachment fold (B-B') along the trend of the major hanging wall paleovalley causes truncation of the older onlapping, progressively folded conglomerates.See Figure

Figure 9b .[Figure 10 .Figure 11 .
Figure 9b.Rapid lateral facies changes in the major paleovalley are displayed by the abrupt thinning and fining toward the southwest of the conglomerate beds.In contrast, the "passing conglomerate" and the overlying basal coarse-grained bed of unit 3 taper toward the northeast, probably due to syndepositional uplift along the Montsec thrust.Location of section B (Figure 10) is shown.A cone of talus breccia occurs along the flank of the Cretaceous megaclast on the left.
the study area was deposited above unit 2 from the north limb of the Oliana anticline to the trace of the Montsec thrust.It also covers with little deformation the crest of the detachment fold.Deposition of unit 3 appears to coincide with motion on the Montsec thrust, which placed unit 3 in a footwall position with respect to the active thrust.The depositional and deformational history of unit 3 and its contrasts with the underlying units are influenced by its structural position and the way in which deformation was propagated through the footwall.Sedimentology.Most of the exposures of unit 3 occur in steep, mostly inaccessible cliffs for which we have no detailed sedimentology.Nonetheless, some striking contrasts with unit 2 are apparent.Most importantly, rather than deriving solely from local source areas, the conglomerate clasts indicate significant derivation from the Axial Zone.The inference that these should then belong to longitudinally flowing rivers is borne out by palcocurrents in the basal parts of unit 3 (Figure 10) which clearly show transport to the SW, parallel to the structural front.Thus, in contrast to the sloping transverse surfaces in the palcovalleys of unit 2, the depositional surface of this unit should be essentially horizontal in sections transverse to structural front.The conglomerates of unit 3 are clast-supported, display broad bedding (1-3 m), and contain uncommon, but well-bedded, fine sandstone interbeds.The clasts of unit 3 are generally rounded to subrounded, suggesting they are farther traveled than most of the clasts in the proximal paleovalleys of unit 2. In areas close to the Montsec thrust, the successions of conglomerates and intervening fine-grained siltstones in unit 3 display more tabular geometries and cover broader areas than do the underlying strata within the paleovalleys.The total stratigraphic thickness of unit 3 thins southward toward the Oliana anticline, and the generally even proportions of very coarse-and fine-grained depositional units in proximal areas are replaced by increasingly dominant conglomeratic strata toward the south.
Because it occurs near the base of unit 3 at a time when renewed motion on the Montsec thrust was just beginning[Burbank et al., 1992], it appears likely that prior to this latest phase of deformation, there were at least several hundred meters of residual topography on the Montsec hanging wall as a result of middle and early Eocene motion on this thrust [Vergds and Mu•qoz, 1990].Around the megaclast's margins, talus cones developed, and later, both these cones and the megaclast were passively enveloped by the aggrading beds of unit 3. Age.Unit 3 has not been directly dated in the study area.present study area to the northwest of the anticline.This yields a mean rate of-15 cm/kyr, which is essentially unchanged from the mean rate of accumulation of unit 2. Geometrical relationships.At the large cross-sectional scale, the thinning and onlap of unit 3 southward on to the north flank Oliana anticline (Figure 3) clearly indicate this large anticline was still growing at this time.In proximal areas adjacent to the Montsec thrust, the strata of unit 3 have been strongly folded (Figure 9) into a footwall syncline.Small thrust faults occur along some of the contacts between the megaclast and adjacent conglomerate beds.During folding, the theological contrasts between the megaclast and the surrounding strata probably caused the more rigid block to glide along minor out-of-syncline thrusts.On the north side of the megaclast, strata of unit 3 are steeply dipping and display a fanning geometry ranging at present from dips of 20 ø overturned to north to dips of 65 ø to the south.This syndepositional wedge-shaped geometry in unit 3 is interpreted to result from folding and uplift of the Montsec footwall.Gentle warping of unit 3 as it passes over the crest of the detachment fold suggests that this anticline and the hanging wall syncline to its north (Figures 2 and 4) continued to tighten in response to Montsec thrusting.Directly above the limestone fin that defines the eastern margin of the minor paleovalley (Figure 9a and 9c), the basal conglomeratic strata of unit 3 are folded into an anticline.It appears that the strata of these beds thicken on each flank of the anticline and thin over its crest, whereas the underlying siltstones and paleosols clearly thicken along the fold axis.the study area and is preserved in steep cliffs that are generally inaccessible.On all sides of the Oliana anticline, unit 4 is extensively preserved and represents the last major depositional event in this region.The magnetic data [Burbank et al., 1992] indicate that initial deposition of unit 4 began at -34.0 Ma.When sedimentaccumulation rates from the underlying strata are extrapolated to the top of this unit, they suggest that it may span 3-4 m.y.[Vergds and Burbank, 1994].Sedimentology.This unit contains abundant clasts that were derived from the Axial Zone and consists almost entirely of clastsupported, rounded, and often well-imbricated conglomerates.Most of the coarse-grained depositional units are 1-3 m thick and display a normally graded to ungraded character.In contrast to the underlying units, there are very few fine-grained intervals preserved within unit 4 (<_5%).Where there are thin units of intervening siltstones, however, long gutters are very common in the base of the overlying conglomerates.In combination with clast imbrications, these gutters provide unambiguous paleoflow directions toward the SW.Therefore unit 4 was deposited from rivers flowing parallel to the structural grain and carrying fartraveled clasts into the foreland.Where unit 4 is preserved directly adjacent to the Montsec hanging wall, small, steeply inclined cones of breccia abut the lower part of the Montsec escarpment.There is no evidence that unit 4 overtopped the Montsec hanging wall in this transect.In fact, the topographic relief on the Montsec hanging wall was much greater than the thickness of unit 4, which was represented by a steep apron of talus breccias banked against the lower slopes of the hanging wall.Although the total amount of topography inherited from middle and early Eocene deformation is unknown, it is likely to represent about half of the present relief, given the limited amount of late Eocene Montsec shortening [Vergds and Burbank, 1994].Geometrical relationships.Because it is the stratigraphically and topographically highest syntectonic unit, unit 4 is more dissected than the older depositional units, and its complete geometry is more difficult to describe.Nonetheless, its proximal strata are only very slightly deformed by the Montsec thrust which gently folds them in several places.This absence of strong deformation indicates that major motion on the Montsec had terminated when deposition of unit 4 commenced.Unit 4 thickens away from the Montsec hanging wall toward the south and SE (Figure 2) and then thins as it approaches the backlimb of the Oliana anticline.Our tectonic reconstructions suggest that most of the thrusting and folding within the anticline itself ceased at this time [Burbank et al., 1992].Nonetheless, the thinning of unit 4 adjacent to the Oliana anticline indicates that some topography persisted across its structural crest, where today erosion of the weakly resistant marls in the core of the fold have caused a strong topographic inversion.Discussion The growth of the Oliana anticline and the coeval defonnation of structures related to the Serres Marginals and Montsec thrust sheets represent the fundamental tectonic controls on both deposition and topography in the study area.Today, the Segre River cuts directly through the strutrurally highest part of the Oliana anticline.During deformation, however, this structure presented an oblique barrier to flow of the ancestral Segre River and caused it to divert to either flank of the growing fold (Figure 12).Along the southeastern limb of the Oliana anticline, low topography persisted southwards into the undeformed foreland, and paleocurrent indicators [Burbank et al., 1992] suggest a fanning of flow directions once the fiver passed the northeastern nose of the anticfine, which controlled the entry point of the river into the foreland.On the northwestern limb of the anticline where we have focused the present study, the ancestral Segre was channeled along the southwest trending structural moat, represented by an irregular synform between the Montsec hanging wall highlands and the Oliana anticline (Figure 2).The progressively fanning dips and the tapered geometry of strata within the growth syncline of unit 2 provide clear evidence of syndepositional rotation of the forelimb of a growing detachment fold.During early stages of deposition and fold growth, rapid accumulation primarily promoted onlap of inherited topography.The rates of vertical uplift are interpreted to have been slower than the accumulation rates (-25 cm/kyr) at this stage.During the succeeding stage of deposition (36.4-35.7 Ma), accumulation rates in the growth syncline were slow (-6 c:n/kyr), finer grained strata were deposited, and the paleovalleys attained their maximum depths as they incised across the crest of the detachment fold.Aggradation in the growth syncline accelerated to ~14 cm/kyr between 35.7 and 35.0 Ma.Continuous tapering of strata toward the syncline's margin, combined with the absence of clear evidence for either strong overlap or offlap with respect to the fold crest, suggests that vertical rates of flank uplift and fluvial deposition were nearly equivalent along the forelimb of the detachment fold.Thus during deposition of unit 2, the longitudinally flowing river was confined between the Oliana anticline and the detachment fold.Despite the ongoing uplift at rates of 0.1-0.3mm/yr in the folds defining the margins of this trough, no significant topography was created within the depositional plain.The positioning of the paleovalleys provides insights on the roles played by lithology and structure in controlling incision, whereas the infilling of the paleovalley reveals aspects of the topography, deformation, and source areas in the Serres Marginals hanging wall.Not unexpectedly, folding and uplift of variably resistant bedrock strata lead to accentuated erosion of weaker beds.The minor paleovalley is enclosed by a resistant limestone and is localized along the axis of a sandstone-and marl-cored fold.These same resistant limestones also delineate the highest preserved point along the only preserved margin of the major paleovalley, which is incised into less resistant red beds.

Figure 12 .
Figure 12.Comparison of the major structural and depositional elements in the vicinity (a) of the Oliana anticline and (b) of Wheeler Ridge, southern California.The absence of a major fiver at Wheeler Ridge, combined with more rapid uplift of the anticline and impinging thrusts, control the large-scale differences in the depositional geometries between these sites.The transverse valleys and smaller-scale drainage patterns at Wheeler Ridge provide reasonable analogs for the palcovalleys in Setres Marginals anticline.river.It appears that detrital sediment production fxom the small, steeply inclined, and largely limestone-dominated local catchments was overwhelmed by the contributions of the axial rivers.The incision of the larger transverse palcovalley through the crest of the growing Serres Marginals anticline caused a locally important unconformity to be generated within unit 2 in the growth syncline.This resulted in recognizable along-strike variations in the geometric relationship between syntectonic strata and the folded pre-Eocene strata.The deeper incision that occurred along the major palcovalley caused juxtaposition of unit ).In contrast, much of the topographic relief of the Montsec hanging wall is likely to have been inherited.The size of the megaclast within the basal part of unit 3 suggests a minimum of several hundred meters of hanging wall relief during early reactivation of the Montsec thrust.The inclined talus cones of unit 4 that are aligned along the lower part of the Montsec hanging wall indicate that a minimum of 500 m of relief above the depositional plain existed at the end of thrusting, during which time -1 km of shortening and nearly equal amounts of hanging wall uplift took place.Sequential reconstructions across the deforming Serres Marginals and Montsec thrust sheets (Figure 2) indicate that between 35 Ma and 33 Ma, -750 m of shortening (27%) occurred between the pinning points on the Montsec thrust and the innermost Serres Marginals thrust [Verggs, 1993].After-35.6Ma,the Montsec footwall was differentially tilted and uplifted, such that the proximal regions were raised 250 m with respect to coeval, formerly horizontal surfaces only 1 km away.This uplift was accomplished primarily through folding.Both preexisting folds in the pretectonic strata and newly deposited strata of units 2 and 3 were tightened into chevron geometries displaying -60 ø interlimb angles (Figure2)[Ramsay, 1974].This late stage uplift supports a model suggesting that during late stages of deformation, horizontal shortening is often superseded by vertical uplift driven primarily by folding.This phenomenon can generate considerable topographic relief late in the deformational sequence.At Oliana, this thickening and uplift shift the longitudinally flowing river toward the foreland (Figure2), where it impinges on the northern flank of the Oliana anticline.
Depositional and topographic patterns atOliana developed in response to thrusting, growth of a detachment fold, anticlinal uplift, and continued deformation of previously folded structures during late Eocene times.Our analysis shows the following: 1.Along the deformational front, there exists an interplay between longitudinally flowing rivers from distant sources and transversely flowing rivers of local origin.In regions where the eroding hanging walls comprise resistant strata and are characterized by relatively small catchments, the transverse contribution to the sedimentary record will be diminished.Where there is high topography on hanging walls inherited from earlier deformation, however, large landslide blocks can tumble into footwall synclines.2. Folding associated with both the forelimbs and backlimbs of growing structures causes related thinning and deformation of syntectonic strata.The resulting wedge-shaped geometries of syntectonic strata open both toward the foreland and the hinterland and record fold growth.3. When folding and crestal uplift are rapid in comparison to the rate of aggradation, relief of the hanging wall above the depositional plain increases, transverse valleys become accentuated, strong offlap occurs in depositional zones, and proximal unconformities may develop as older syntectonic strata are folded, uplifted, and exposed to erosion.4. Transverse paleovalleys are localized by the structure and lithology into which they are incised.The cores of folds and weakly resistant strata provide favored pathways.At Oliana, resistant limestones uphold the valley walls, and synclinal folds have been exploited by the paleovalleys.When crestal uplift rates are only a few millimeters per year, paleovalley incision can sometimes keep pace with uplift.The topographic relief on such paleovalleys at Oliana is > 150 m.The transition from incision to deposition within a paleovalley is apparently controlled by accelerating rates of aggradation in comparison to hanging wall uplift.During filling of the transverse paleovalleys, local source areas are almost exclusively represented.