The rise of a mountain range affects moisture circulation in the atmosphere
and water runoff across the land surface, modifying the distribution of
precipitation and drainage patterns in its vicinity. Water routing in turn
affects erosion on hillslopes and incision in river channels on surrounding mountain ranges. In central Guatemala, two parallel, closely spaced mountain ranges formed during two consecutive pulses of uplift, the first between 12 and 7 Ma (Sierra de Chuacús–Sierra de las Minas), and the second after 7 Ma (Altos de Cuchumatanes). We explore the climatic and tectonic processes through which the rise of the most recent range drove the slowing of river incision and hillslope erosion over the previously uplifted range. The
A mountain range affects both the circulation of atmospheric moisture around its relief and the flow of precipitated water over the land surface on its slope and in the vicinity. Moisture rises and precipitates on its windward side, while rain shadows tend to be cast over its lee-side and over the land surface located downwind (e.g., Meijers et al., 2018; Galewsky, 2009). Once precipitated, the fraction of water that runs off as overland flow drives hillslope erosion and river incision. River drainages are dynamic systems that can transmit forcing along drainage lines in both the downstream and upstream directions. Changes in climate and vegetation modulate hillslope erosion, and these changes are then transmitted to rivers from the headwaters downstream. Conversely, changes affecting downstream river reaches, such as the adjustment of river gradient to the rise of mountain ranges or to variations in sea level, can be transmitted upstream along river channels (Humphrey and Heller, 1995; Whittaker and Boulton, 2012). From there, they are transmitted uphill along valley slopes (Harvey, 2002; Mudd and Furbish, 2007). Mountain ranges affect the erosion of surrounding reliefs through this combination of top-down and bottom-up processes.
The growth of contractional orogens commonly involves the outward, sequential propagation of contraction, and the formation of successive, in-sequence mountain ranges. Moisture, by contrast, is commonly advected in the opposite direction, from the forelands to the orogen interiors. Precipitated water is then commonly returned to the foreland by river networks that flow from the orogen interiors to the forelands. The rise of front ranges therefore commonly takes place both upwind and downstream of preexisting ranges, driving the aridification of previously uplifted ranges (Garcia-Castellanos, 2007). The new front ranges will also affect the rivers that drain the previously uplifted ranges toward the forelands. These rivers will adjust to enhanced rock uplift by enhanced incision through the steepening of their gradient in areas of enhanced uplift (Leland et al., 1998). This steepening is accompanied by a transient decrease in incision rate upstream of the rising ranges (Champel et al., 2002). In some cases, the topography and underlying tectonic structure will respond fast enough to reach a new dynamic equilibrium between relief, climate, and tectonics (Willett and Brandon, 2002; Whipple and Meade, 2006) without any substantial alteration of the drainage network. In other cases, however, the range-transverse river networks will experience reorganization before the equilibrium is reached (Jackson et al., 2002; van der Beek et al., 2002; Brocard et al., 2012). In a few cases, the slowing down of landscape response time, as a result of aridification, upstream and downwind of the rising ranges will even lead to the disintegration of river drainages and to the topographic decay of interior ranges. This evolution characterizes the nucleation and growth of orogenic plateaus formed by lateral accretion (Sobel et al., 2003; Garcia-Castellanos, 2007). These different evolutionary pathways have been explored at the scale of entire orogens. They are, however, seldom documented at the scale of individual mountain ranges because their manifestation at that scale is harder to separate from more local signals driven by spatial variations in bedrock erodibility, stochastic processes (e.g., landslides), and topographic inheritance.
We document here how the rise of a recent mountain range (the
170
New
Shaded topography of the study area, showing the tectonic setting
of central Guatemala
Left-lateral motion along the North American–Caribbean plate boundary in
central Guatemala has produced elongate ranges parallel to the plate boundary
(Fig. 1a). We investigate the growth and erosion of two of these ranges,
namely the Sierra de Chuacús–Sierra de las Minas range (SM-SC range),
and the Altos de Cuchumatanes range (AC range, Fig. 1b). Rocks in the SC-SM
range possess a deeply penetrative, sub-vertical tectonic fabric, imparted by
70 Myr of left-lateral wrenching along the Caribbean–North American plate
boundary (Ratschbacher et al., 2009; Ortega-Gutierrez et al., 2004;
Ortega-Obregón et al., 2008). Since Eocene time, left-lateral motion has
been accommodated by the Motagua fault and, to a lesser extent, by the
Polochic fault (Fig. 2b). The Motagua fault is, with
Age of geomorphic markers and drainage lines in central Guatemala. Ages of Miocene valleys and Quaternary paleovalleys (V1–V12) in Myr. Data source are as follows: 1: Brocard et al. (2011); 2: Brocard et al. (2012); 3: Tobisch (1986); 4: Plio-Quaternary lacustrine deposits (Brocard et al., 2015a). Newly dated lavas are defined as follows: CJY: Chujuyúb; EJ: El Jute. Range names are defined as follows: AC: Altos de Cuchumatanes; SC: Sierra de Chuacús; SM: Sierra de las Minas; VA: volcanic arc. Faults are defined as follows: MF: Motagua; PF: Polochic. The background used is taken from the shaded GTOPO 30 DEM.
Incision below the middle Miocene Maya surface, based on the
elevation of surface remnants (upland relict surface). Incision contour line
spacing is 1
Today, central Guatemala is straddled by 3–4
Uplift propagated north of the Polochic fault during the late Miocene (Brocard
et al., 2011) and was marked by the rise of the AC range in response to
contraction within the North American plate (Authemayou et al., 2011). The
rise of the AC range drove widespread reorganization among the rivers that
drain the northern flank of the SC-SM range (Fig. 2). Numerous river valleys
were then abandoned and left stranded on the rising AC range. Their
deformation indicates that the AC range has risen
Mean annual precipitation across the study (MARN, 2016) from
dry (red) to wet (blue). Isohyet spacing is 100
While contraction has defined the evolution of the western part of the study
area, transtension has prevailed further east since the late Miocene at
least. The dominance of transtension in the east results chiefly from an
eastward increase in the divergence angle between the strike of the plate
boundary and the direction of plate motion (Rogers and Mann, 2007).
Transtension led to growth of the Lake Izabal basin (Fig. 1a), which is has
been filled with
The rivers located along the northern flank of the SC range represent the headwaters of a network that, farther downstream, experienced widespread reorganization during the late Miocene (Fig. 2). Reorganization led to the formation of range-parallel rivers halfway between the drainage divides of the SC-SM and the AC ranges. These E–W-striking rivers collect rivers that flow north, down the northern flank of the SC-SM range, and funnel them into the Chixóy River, one of the few rivers, together with the Cahabón, Chixóy, Selegua, and Cuilco rivers (Fig. 2), that still cross the AC range. These latter rivers also straddle the trace of left-lateral Polochic fault before entering the AC range. The Polochic fault has deflected and lengthened their course (Fig. 2) since the late Miocene (Brocard et al., 2011). Plio-Quaternary transtensional faulting along the northern flank of the SM range initiated a second (and still ongoing) pulse of drainage reorganization during the Quaternary (Fig. 2) (Brocard et al., 2012).
Since the Pliocene, large volcanoclastic aprons have piled up along the NE side of the Central American Volcanic arc. They have buried the western end of the SC range, deranging its river network (Fig. 2). This complex area is therefore excluded from the present study. Likewise, the karstic highlands of central Guatemala, especially those located north of the Polochic fault, are also excluded from the analysis because their dynamics is strongly influenced by the high-frequency opening and closure of subterranean karstic pathways (Brocard et al., 2015a, 2016a).
Moisture tracking from the Pacific Ocean and from the Caribbean Sea is
intercepted by slopes that face the western and eastern coasts of Guatemala
and the Petén lowlands in the north (Fig. 4). The AC range receives
4–6
Rock belts in central Guatemala tend to follow the strike of the mountain ranges (Fig. 5). Late Cretaceous schists and gneisses of the Chuacús Formation (henceforth, formation will be abbreviated to Fm.). form the core of the SC-SM range. They are flanked by the late Cretaceous migmatites of the San Agustin Fm., and by marbles and amphibolites of the Jones Fm. In the north, this metamorphic core is tectonically juxtaposed, across the Baja Verapaz shear zone, to the basement of North America, which is covered mostly by a Permian megasequence of terrigenous sediments and carbonates (Sacapulas, Tactic-Esperanza, and Chochal Fms.; Anderson et al., 1973). The basement and cover are intruded by Paleozoic (e.g., Rabinal), Triassic, and Jurassic (e.g., Matanzas) plutons.
Geology and structure of central Guatemala (Instituto Geográfico Nacional de Guatemala; Instituto Hondureño de Geología y Minas; Instituto Nacional de estadística y geografía de México), draped over the GTOPO 30 DEM. AC stands for Altos de Cuchumatanes, IXF stands for Ixcán fault, MF stands for Motagua fault, PF stands for Polochic fault, SC stands for Sierra de Chuacús, SM stands for Sierra de las Minas, and VA stands for volcanic arc.
A megasequence of continental terrigenous sediments (the Todos Santos Fm.), Cretaceous carbonates, and Cretaceous evaporites (Cobán Fm., Campur Fm.) covers much of the AC range (Fig. 5). Ultramafic rocks obducted over the carbonates in late Cretaceous (Campanian) time (Fourcade et al., 1994) are preserved within weakly metamorphic synformal klippes (Baja Verapaz, Santa Cruz, and Juan de Paz ophiolites). Higher-grade serpentine mélanges crop out along the Motagua valley (Flores et al., 2013).
The southern base of the SC-SM range is incised into sediments deposited in
narrow transtensional basins along the Motagua fault (Ratschbacher et al.,
2009). During Eocene times (Newcomb, 1975), one such basin was filled by
continental red beds of the Subinal Fm. (Fig. 5), which has an exposed
thickness of
The age of the low-relief Maya surface was previously constrained by
bracketing age markers, such as 14–15
We measured the concentration of
Catchments sampled for the
The majority of the samples consist of river-borne quartz collected in 30
rivers that drain the SM, SC, and AC ranges (Fig. 6, Table S2-2, Fig. S2-3 in the Supplement). They provide catchment-averaged hillslope erosion rates (Brown et al., 1995). Quartz was extracted from the sand grain-size fraction (250–500
To test the influence of hillslope steepness and of precipitation on
We extracted the long-profiles of 220 rivers located in the AC, SC, and SM
ranges, using the Guatemala national 20
Distribution and grouping of the streams used in the river long-profile analysis, and their grouping by geographic areas. AC stands for Altos de Cuchumatanes, LI stands for Lake Izabal, PF stands for Polochic fault, SC stands for Sierra de Chuacús, and SM stands for Sierra de las Minas. Numbers (1–106) correspond to the numbers ascribed to the rivers, as listed in table S4-1. The corresponding river profiles are presented in Figs. S4-2 to S4-7 in the Supplement and in Fig. 12. Boxes A–D show footprints of the maps displayed on Fig. 8.
We resorted to a linearization method that filters out the downstream increase
in stream discharge to identify knickpoints along river profiles. We chose
the integral method (Perron and Royden, 2013), in which elevation is plotted
(on chi plots or
Streambed morphology was examined along each linearized segment using
stereoscopic 0.5
Convex-up breaks in slopes along river profiles are commonly referred to as river knickpoints. For convenience, we refer here to all breaks in slope, whether convex or concave, as knickpoints. Knickpoints were classified as lithogenic, alluvial, tectonic, migrating, and miscellaneous (see Supplement 4). Miscellaneous knickpoints represent adaptations of river profiles to local, stochastic disturbances (such as landslides and epigenies) and are usually short-lived. Most knickpoints in the study area reflect adaptations of river gradients to along-stream variations in rock uplift rate, bedrock erodibility, sediment flux, or sediment grain size. These knickpoints can be regarded as steady, inasmuch as their location only changes very slowly along the river profiles, tracking spatial changes in the distribution of rock types, rock uplift, sediment fluxes, and bed load grain size. By contrast, knickpoints that spearhead step increases or step decreases in river incision rates migrate in the upstream direction along river profiles in the form of waves of accelerated (Rosenbloom and Anderson, 1994; Merritts et al., 1994) or decelerated (Howard, 1997) incision. They are hereafter referred to as migrating knickpoints, for they usually migrate faster than the knickpoints previously described. Concave-up migrating knickpoints commonly mark the transition from detachment-limited to transport-limited river incision (Whipple and Tucker, 2002) and are usually found at the apex of alluvial fans (Fig. 8c) and pediments (Fig. 8b).
Examples of some knickpoint types presented in their geomorphic
setting. Shaded and sloped 30
Theoretical geometric differences between migrating and steady knickpoints in
linearized spaces have been used to discriminate unstable, migrating
knickpoints from stable, equilibrium knickpoints (Goldrick and Bishop, 1995;
Perron and Royden, 2013; Whipple and Tucker, 2002). In
Additional discriminating elements must be used. A first screening consisted
of checking whether the knickpoints coincide with marked variations in bedrock
erodibility, rock uplift rates, or local anomalies, in which case they were
classified as steady. To assess the effect of lithological variations we used
The method has some limitations: first, local variations in bedrock erodibility maybe not be systematically detected, as a result of the imprecision of geologic maps, especially in the least accessible parts of the SM and AC ranges. Second, large intra-formational changes in facies can generate variations in bedrock resistance as sharp as (or even sharper than) erodibility differences between mapped geological units. These two effects may lead to the interpretation of stable knickpoints as migrating knickpoints. Conversely, some migrating knickpoints may be pinned to lithological contacts (Crosby and Whipple, 2006), and filtered out by the analysis. Nonetheless, we consider that, given the large number of analyzed knickpoints, the analysis captures the most import aspects of the evolution of the landscape within the study area.
The Maya surface (Figs. 2 and 3) likely formed close to sea level because it
can be traced to the coast of the Caribbean Sea (Brocard et al., 2011). It was once covered by extensive fluvial deposits, especially south of the Motagua fault, where the fluvial deposits are preserved below extensive ignimbrites (Williams and McBirney, 1969). The lahar deposit of Chujuyúb rests directly on a thick saprolite that blankets the Maya surface. Lahar
emplacement predates the incision of a 450
Evolution of incision rates in the studied ranges.
Letters a–j correspond to river incision rates inferred from
The chronology of incision along the southern side of the SC range is
documented by remnants of basalt flows scattered along the floor of the
Motagua valley. These flows track from vents located south of the valley on
the Caribbean plate (Tobisch, 1986). The outcrop of El Jute represents the
distal end of a lava flow that abutted the base of the SM range, backfilling
the Huijo River valley with
The chronology of incision can be refined by incorporating the previously
dated basalts (Tobisch, 1986). The closest occurrence, located 6
Incision of the Motagua valley, from the elevation of the basalt of El Jute
down to the current valley floor, would thus have occurred between 6.1 and
3.1 Ma at
Large steeply dipping faults bound the Eocene fill of the Motagua valley. Dip-slip on these faults could be responsible, in part or in whole, for the deepening of the Motagua valley, a possibility contemplated by Tobisch (1986). Various traits of the valley, however, rule out any substantial contribution of these faults. First, fluvial sediments have bypassed the Motagua valley since Eocene time, feeding a transtensional basin at the lowest eastern end of the Motagua valley (Fig. 1b). Second, the alluvial fans that have grown astride these faults show no evidence of faulting, nor any anomaly in their catchment/fan surface ratios (Tobisch, 1986). Third, the faults encountered along the base of the SC-SM range exhibit only ancient, ductile to ductile-brittle left-lateral deformation (Bosc, 1971; Roper, 1978). Finally, the middle Miocene low-relief surfaces lie at about the same elevation north and south of the Motagua fault (Simon-Labric et al., 2013). Extension of antithetic boundary faults would need to remain well-balanced, despite hundreds of kilometers of left-lateral displacement along the Motagua fault since the middle Miocene, to avoid the development of significant offsets of these surfaces. The deepening of the Motagua valley therefore appears to have been achieved by erosion, through the removal of the erodible Eocene sediments that filled the Eocene fault basin, giving the valley the appearance of a recently active graben.
The incision chronology of the AC range is constrained by transverse
paleovalleys that are shallowly incised into the Maya surface (e.g., Figs. 8a, i and 9). Uplift of the AC range since their abandonment provided space for the incision of 1500–2600
Variations in detrital
Catchment-averaged detrital
In the AC range, erosion rates (arrows, Fig. 10) show a marked increase from the drier and less steep highlands to the wet and steep frontal slopes (from CATA to CHEL to XAC). The SM range displays a similar trend of increasing erosion down the mountain flank, as entrenchment in the Maya surface increases (from COL to FRI to RAN), with one outlier (SLO). The magnitude of increase in the SM range is intermediate between the ones measured in the SC and CA ranges. In the SC range indeed, a downstream increase would be expected to occur first in the downstream direction, between the drainage divide and the mountain flanks, as a result of the decreasing contribution of slowly eroding low-relief uplands with downstream distance (Willenbring et al., 2013b). Increase should be followed by a decrease in erosion rate as rivers start draining pediments that floor the Chixóy River catchment. An increase in erosion rate, downstream of the paleosurface, is measured (from PAS to PAE), but it is much less pronounced than in the AC range. The decrease in erosion rate is also very subdued (from XEU to CUB). In one case (from SMS to SMM to SMI), no increase or decrease is observed.
A total of 9
Distribution of rivers according to the number of segments identified in each river.
The distribution of alluvial reaches is bimodal in the SC-SM range (Fig. 12a1 and b2): alluvial reaches tend to be found either at the base of the mountains or at high elevation over the remnants of the Maya surface (e.g., Fig. 8a and d). High-elevation alluvial reaches tend to transport a rather fine-grained bed load, composed of sand derived from the weathering of micaschist, gneiss and granite, and gravel derived from quartzose veins and silicified pegmatites (Brocard et al., 2012). Intermediate-elevation alluvial reaches occur upstream of obstructions, most notably landslides in the SM-SC range, over extremely erodible fault damage zones, and within localized areas of tectonic subsidence (especially along the Polochic fault corridor, on the southern flank of the AC range).
Boulder reaches are found mostly on crystalline rocks. There, they are more frequent on the wet slopes SM range than on the dry slopes of the SC range. In the SM range, many boulder-strewn reaches form after the winnowing of the fine-grained matrix of debris flows. The SM range is the first range hit by tropical depressions tracking from the Caribbean Sea. They trigger numerous landslides along the wettest slopes of the SM range (Ramos Scharrón et al., 2012; Bucknam et al., 2001). Because SM range soils are more frequently close to water saturation, they are more likely to be affected by landslides when earthquakes strike the range (Harp et al., 1981). In the SM range, boulder armoring is common on the serpentinite mélanges that crop out up to high elevations along its southern flank, owing to the presence of knockers in the mélanges (e.g., Fig. 8c). Boulder-strewn reaches in the AC range form over phyllites. There, boulders are made of the most resistant beds of Pennsylvanian phyllites and of sandstone and limestone blocks derived from overlying formations that slid along valley flanks down to the streambeds.
Bedrock river reaches are most commonly found downstream of convex migrating knickpoints, the distribution of which is presented in the following section.
Among the 350 identified knickpoints, 40
Details about the significance of the distribution of steady knickpoints, as well as a more systematic review of the origin of
all identified clusters of migrating knickpoints, are
provided in Supplement 4. Some migrating knickpoints
can be tied to well-identified and well-dated river Quaternary diversions
(e.g., S3-1 to S3-3, Fig. 12a2, Brocard et al., 2012). Most migrating
knickpoints dot the brim of upland low-relief surface remnants (Figs. 12a1 and
a2 and 8d). They may have therefore initiated when the Maya surface started
being incised, at
The decline of river incision rates in the SC-SM range was coeval to the rise of in the AC range (Fig. 9). This can reflect a complete transfer of rock uplift from the SC-SM range to the AC range, but it remains surprising that river incision rates declined so sharply within the SM range, considering that the range had not undergone any substantial topographic decay. Likewise, very low hillslope erosion rates are maintained on steep slopes within the SC-SM range today. A genetic relationship can therefore exist between the rise of the AC range and the decline of river incision and hillslope erosion in the SC-SM range. The rise of the AC range may have affected incision rates in the SC-SM range in two ways. First, by decreasing moisture delivery to the SC-SM range, it may have reduced hillslope erosion rates and the delivery of water and sediment to the streams, thereby decreasing river incision rates. Second, by forcing the drainage of the northern side of the SC range to adjust to rock uplift in the AC range, it promoted a decrease in river incision rates upstream of the AC among the rivers of the SC range that still cross the AC range. After reviewing the respective contributions of these top-down and bottom-up processes, we analyze how they combined to affect the present-day morphological evolution of the SC-SM range.
Silicate weathering is 3 times faster on the wet
(1800–3000
For MAP
It seems therefore that the most arid parts of these ranges erode slowly to
very slowly, despite maintaining steep slopes. Besides, slope steepness and
MAP become poor predictors of short-term (10
The sequential rise of the SC-SM and AC range generated an evolving pattern of hillslope steepness and of precipitation that may have contributed to the decline of incision in the SC range since the middle Miocene. We review hereafter evidence for changes in climate and tectonics susceptible to having impacted hillslope steepness and erosion over time.
The distribution of precipitation (Fig. 2) shows that the AC range currently
prevents the ingress of moisture tracking from the Yucatán and Petén
lowlands toward the SC range. Moisture was therefore likely able to reach the
SC range before the rise of the AC range. Such a deep penetration of
moisture is supported by paleo-precipitation estimates obtained from the
analysis of tree species and of paleosols in the 7.4 Ma forest of Sicaché
(Fig. 2). This subtropical forest was growing on the floor of a late Miocene
paleovalley. The geochemical characteristics of its paleosols suggest mean
annual precipitation between 950 and 1300
The rise of the AC range sparked widespread reorganization of the range-transverse drainage (Brocard et al., 2011), tectonically defeating many rivers that used to cross the AC range. Some rivers maintained a course transverse to the rising structure by steepening their gradient. We review hereafter the evidence for faster rock uplift in the AC range than in the SC range, river course lengthening along the Polochic fault, and decreased rock uplift rates in the SC range. We review the contribution of each of these tectonic processes to the decline of river incision in the SC range.
The deformation of paleovalleys abandoned during the uplift of the AC range
documents
Box plots of stream segment normalized steepness as a function of
streambed type and location.
The steepness of river profiles in the AC range further support the hypothesis
that the range still rises faster than the SC range. The projection of river
profiles in
Boulder-armored and alluvial channels are also steeper in the AC range (Fig. 14b and c). Boulders act as bedrock, and boulder-armored channels can therefore be expected to behave like detachment-limited channels. Alluvial channels are likely transport-limited, and therefore their gradient is less sensitive to rock uplift (Whipple and Tucker, 2002; Cowie et al., 2008). The observed increase in alluvial channel gradient therefore most likely reflects an increase in bed load grain size with increasing erosion rate, resulting from shorter residence time and limited comminution of bedrock blocks in hillslope soils (Riebe et al., 2015; Neely and DiBiase, 2020).
The four rivers that still cross the AC range are the Cuilco, Selegua,
Chixóy, and Cahabón rivers (Fig. 7). Of these, the Cahabón River
is the smallest. It is affected by an ongoing pulse of drainage rearrangement,
which over the past
Long profiles of rivers transverse to the AC range (Chixóy, Selegua, and Cuilco), with indications of reaches affected by vertical rock uplift (grey area), and those affected by horizontal lengthening along the Polochic fault (hatched area). S and L represent contributions of steepening and lengthening to the uplift of river profiles upstream of the AC range. River knickpoint nomenclature is listed in Fig. 12.
The steepening of river profiles across the AC range occurs in response to the
rise of the AC range in an area previously characterized by a foreland through
which flowed shallow-gradient rivers. The phase of steepening that converted
these shallow-gradient rivers to steep-gradient transverse rivers should have
ended when the gradient of the transverse rivers became steep enough for river
incision to counterbalance rock uplift in the AC range. Steepening requires
the headwaters to rise with respect to the foreland. Such rise requires a
transient imbalance, upstream of the AC range, during which incision rates are
lower than rock uplift rates. The sharp drop in incision rates in the SC range
from 145–205 to
Using the river profile of the Chixóy River (Fig. 15), it can be assessed
that
River courses are progressively offset by felt-lateral slip on the Polochic
fault, leading to the development of ever-lengthening tectonic deflections
above the fault. Maintaining gradients along these deflections sufficiently
steep to allow the downstream transport of the bed load requires uplift of
river channels upstream of the tectonic deflections, commensurate to the
amount of river lengthening. Before the rise of the AC range, left-lateral
deflections formed at the contact between the SC range and the northern
foreland. Streambed uplift along the deflections happened in an area where
streambeds were shallowly incised, allowing frequent avulsions toward the
foreland that annealed deflections (Sieh and Jahns, 1984) and limited river
lengthening. The rise of the AC range promoted deeper entrenchment of the
tectonic deflections along the Polochic fault, preventing the annealing of
deflections and converting slip on the Polochic fault into permanent,
cumulative river lengthening. The courses of the Chixóy, Selegua, and
Chilco rivers have thus been lengthened by 25 to 40
With estimated surface uplift rates of
A decrease in rock uplift rates in the SC range is documented by the
difference in elevation between the paleovalleys located in the AC range, such
as V3 (Figs. 2 and 8a), which underwent 2.8
Between 50 and 300
The absence of incision in the SC range could also be viewed as a lack of response to tectonic forcing, in a landscape subjected to aridification since 7 Ma. By decreasing water discharge, aridification would reduce stream power. Besides, aridification could alter the balance between water and sediment discharges, such as to reduce sediment transport capacity and river incision (Beaumont et al., 1992). It could also dampen the delivery of tools to erode the bedrock (Sklar and Dietrich, 2006). A contribution of climate to the slowing down of incision is further supported by the evolution of the southern flank of the SC and SM ranges. These flanks, albeit exposed to a different tectonic forcing than the northern side of the SC range, underwent a similar decline in incision. The only possible contribution of bottom-up processes to the slowing down of incision along these flanks is the lengthening of the Motagua River at its downstream end into the Caribbean Sea. Such lengthening would promote surface uplift farther upstream in order to maintain the downstream dispersal of sediments. Some lengthening probably occurred during the emergence, at the downstream end of the river, of a transtensional basin (Fig. 1b) that has been filled with terrigenous sediments since the Mio-Pliocene (Carballo-Hernandez et al., 1988). This lengthening of
Modern
Three clusters of migrating knickpoints are found at different elevations
along the northern flank of the SC range (Figs. 8b and 12a2). The uppermost
cluster consists of convex knickpoints that dissect the middle Miocene Maya
surface. They represent the front of an erosion wave that formed in response
to the initial uplift of the SC range at 12 Ma. These knickpoints may have
nucleated near the base of the range along the Polochic fault, which back
then represented the boundary between the SC range and the northern foreland
(Fig. 16a). The second cluster of knickpoints consists of convex knickpoints
located 500
The
The third cluster is located at the base of the mountain. It consists of
concave-up knickpoints dotting the apex of pediments that have formed along
the range (Figs. 8b and 12a2). The pediments are currently extensively buried
under pumice, deposited during a large late Pleistocene eruption (Brocard and
Morán, 2014; Rose et al., 1987). Pediments usually form under semi-arid
climates, along drainages with stable base levels. They grow by extending
their apex into the range (Pelletier, 2010; Strudley et al., 2006; Thomas,
1989). The pediments of the SC range have developed in an area that has been
incising slowly over the past 7 Myr (
The steepest reaches along the northern flank of the SC range are located between the intermediate cluster of convex-up knickpoints, and the basal cluster of concave-up knickpoints. The close association of the middle and lower clusters may not be coincidental. An alternate hypothesis for the formation of the intermediate cluster is that it did not nucleate, as hypothesized above, at the front of the range but on site by faster backwearing of the lower slopes than downwearing of the upper slopes, immediately ahead of the lower cluster. They would have since grown in height and steepness as the pediments extend into the range. The intermediate cluster and basal cluster can be viewed as the lips and toes of large knickzones (or knickpoint faces; Gardner, 1983). Models predict that if water discharge has a larger influence than river gradient on stream incision, such knickpoint faces tend to steepen and amplify over time during backwearing (Weissel and Seidl, 1998; Tucker and Whipple, 2002), which is consistent with the observed topography. The intermediate and basal clusters are not present farther east along the wet flank of the SM range (Figs. 12d and 16c). Only one cluster of large migrating knickpoints separates the very flat uplands of the SM range from its deeply incised, wet lower flank (Fig. 8d). Aridification therefore appears to be responsible for the development of the highly stepped topography of the SC range.
Development of the central ranges of Guatemala. Along profiles A-A* and B-B* (see Fig. 13a for location) (AC: Altos de Cuchumatanes; EJ: El Jute; Mo.F: Motagua Fault; Po.F: Polochic Fault; SC: Sierra de Chuacús; SIC: fossil forest of Sicaché; SM: Sierra de las Minas).
The evolution of the study area can be summarized as follows. Orogenesis started with the rise and coeval incision of the SC-SM range, from 12 to 7 Ma (Fig. 16a). After 7 Ma, rock uplift decreased in the SC range and picked up in the AC range. Fast uplift in the AC range led to the tectonic defeat of many rivers which, upon exiting the SC range, flowed across the foreland. Most of the defeated rivers were rerouted into the drainage of the Chixóy River, one the four rivers that maintained a course across the AC range. Transient steepening of rivers in response to the fast rise of the AC range and the lengthening of these same rivers by the Polochic fault promoted surface uplift within the SC range. The complete cessation of river incision along the northern side of the range may therefore result from a combination of river steepening, river lengthening, and decrease in rock uplift in the SC range. However, the fact the stalling of incision is as complete along the southern side of the range as along the northern side suggests that tectonics was not the only cause but that the aridification of the SC range, resulting from the rise of the AC range, was also instrumental in the stalling of river incision. The decrease in precipitation over the SC range reduced erosion on its hillslopes, decreased river discharge, and therefore contributed to the stalling of river incision on both sides of the range. Precipitation and erosion concentrated on the northern flanks of the AC and SM ranges, while the SC range became almost passively uplifted.
The high erosion rates and wet slopes that characterize the AC range today are reminiscent of the SC range between 12 and 7 Ma. The AC range became the front range, intercepting the moisture that tracks from the foreland, while the SC became an inner range, upstream of the front range and within its rain shadow. A decrease in rock uplift rate alone cannot account for the sharp decline of erosion in the SC range after 7 Ma because the relief and steepness of the range did not suddenly decrease. Aridification within the rain shadow of the AC range contributed to the overall decline in erosion in the SC range. The SC range displays characteristics of dry orogen interior ranges and of orogenic plateaus. It has entered a stage of progressive topographic decay, marked by the development of pediments. Pediment formation occurs in a context of aridity and very low incision rates. Such low incision rates, driven by aridification in the rain shadow of front ranges and by river steepening across the front ranges, characterize the evolution orogenic plateaus formed by tectonic accretion (Sobel et al., 2003; Garcia-Castellanos, 2007). On orogenic plateaus, this evolution ultimately leads to the disintegration of river drainages, isolating dry interior drainages from the forelands. Along interior drainages incision then further decreases, with the rivers being graded to high-elevation base levels. Reduction in local relief then receives the added contribution of the sediment infilling of the now closed catchments (Sobel et al., 2003). If the drying of the SC range was more severe it could ultimately lead to drainage disintegration.
Continued growth of the pediment could lead to the formation of an
intramontane pediplain (Baulig, 1957), at elevations of 0.9–1.2
Isotopic The deformation of paleovalleys indicates that the AC range experienced
The concentration of detrital terrestrial Precipitation is strongly controlled by topographic obstructions resulting
from the rise of the AC range, which intercepts Caribbean moisture.
Precipitation is high along the northern flanks of the AC range but low
over the SC range, which lies within rain shadows. Fossil vegetation
preserved at the base of the SC range indicates a wetter climate at 7 Ma,
which is when the AC range started to grow.
In this context, the slow current hillslope erosion rates in the SC range
appear to be contributed to in part by the rise of the AC and the
development of rain shadows. It can be hypothesized, therefore, that
hillslope erosion rates in the SC range were higher before the uplift of the
AC range. The rise of the AC range led to the steepening of river profiles along the
rivers that maintained a course across the rising range. The steepening of
river profiles triggered a transient decrease in river incision rates
upstream of the AC range. The lack of resumption of river incision upstream
of the AC range implies either that these rivers have not re-established
equilibrium profiles or that other factors prevent the return of incision. The rise of the AC range led to the entrenchment of range-transverse
rivers along the Polochic strike-slip fault. Fault slip has driven
continuous lengthening of river courses on top of the fault over the past 7 Myr. Lengthening contributes to the rise of river profiles upstream of the
Polochic fault, and therefore to the slowing down of river incision in the
SC range. The difference in elevation between paleovalley in the AC range and the SC
range imply rock uplift rates have declined in the SC range. This decrease
in rock uplift rates may have also contributed to the decline in incision in
the SC range The fact that incision decreased equally on either sides of the SC range,
despite widely different tectonic forcing, implies instead that
aridification of the SC range contributed significantly to the decrease of
river incision over the SC range. In the SC range, hillslope erosion slightly outpaces base level lowering,
implying an overall trend to slow topographic decay, despite continuing
surface uplift. The persistence of middle Miocene low-relief surfaces on
mountain tops in the SC range instead implies no net reduction in the range
height, and a decay that proceeds by backwearing rather than downwearing. The slowing down or erosion rates over the SC range has resulted in a
slowing down and stacking of upstream-migrating knickpoints and erosion
waves over the SC range and of the development of pediments at its base.
All data used in this work are provided in the Supplement.
The supplement related to this article is available online at:
GB contributed to project design, river
The authors declare that they have no conflict of interest.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We thank the department of Geology at the Centro Universitario del Noreste (CUNOR) of the University of San Carlos de Guatemala (USAC) for their dependable support during field work. We thank reviewer Paul Umhoefer and the anonymous reviewer for their insightful comments.
This research has been supported by the College of Science and Engineering, University of Minnesota (grant no. 1003-524-5983), and the Swiss National Foundation (grant no. 200020-120117/1).
This paper was edited by Simon Mudd and reviewed by Paul J. Umhoefer and one anonymous referee.