Shaping of topography by topographically-controlled vegetation in tropical montane rainforest.

Topography is commonly viewed as a passive backdrop on which vegetation grows. Yet, in certain circumstances, a bidirectional feedback may develop between the control of topography and the spatial distribution of vegetation and landform development, because vegetation modulates the erosion of the land surface. Therefore, if reinforcing feedbacks are established between erosion and land cover distribution over timescales relevant to landform development, then the interactions between vegetation and topography may create distinctive landforms, shaped by vegetation. We expose here a strong correlation between the spatial distribution of vegetation, erosion rates, and topography at a characteristic length scale of 102-103m (mesoscale topography) in the Luquillo Experimental forest (LEF) of Puerto Rico. We use high-resolution LiDAR topography to characterize landforms, satellite images to classify the vegetation into forest types, and in-situ produced cosmogenic 10Be in the quartz extracted from soils and stream sediments to document spatial variations in soil erosion. The data document a strong correlation between forest type and topographic position (hilltop vs. valleys), and a correlation between topographic position and 10Be-derived erosion rates over 103-104 years. Erosion is faster in valleys, which are mostly covered by monocot Palm Forest, and slower on surrounding hills mostly covered by the dicot Palo Colorado Forest. Transition from one forest type to the next occurs across a break-in-slope that separates shallowly convex hilltops from deeply concave valleys (coves). The break-in-slope is the consequence of a longer-lasting erosional imbalance whereby coves erode faster than hills over landscape-shaping timescales. Such a deepening of the coves is usually spurred by external drivers, but such drivers are here absent. This implies that cove erosion is driven by a process originating within the coves themselves. We propose that vegetation is the primary driver of this imbalance, soil erosion being faster under Palm forest than under Palo Colorado forest. Concentration of the Palm forest in the deepening coves is reinforced by the better adaptation of Palm trees to the erosive processes that take place in the coves, once these develop steep slopes. At the current rate of landscape development, we find that the imbalance started within the past 0.1-1.5 My. The initiation of the process could correspond to time of settlement of these mountain slopes by the Palm and Palo Colorado forests.

pyBadlands: A framework to simulate sediment transport, landscape dynamics and basin stratigraphic evolution through space and time.

Understanding Earth surface responses in terms of sediment dynamics to climatic variability and tectonics forcing is hindered by limited ability of current models to simulate long-term evolution of sediment transfer and associated morphological changes. This paper presents pyBadlands, an open-source python-based framework which computes over geological time (1) sediment transport from landmasses to coasts, (2) reworking of marine sediments by longshore currents and (3) development of coral reef systems. pyBadlands is cross-platform, distributed under the GPLv3 license and available on GitHub (http://github.com/badlands-model). Here, we describe the underlying physical assumptions behind the simulated processes and the main options already available in the numerical framework. Along with the source code, a list of hands-on examples is provided that illustrates the model capabilities. In addition, pre and post-processing classes have been built and are accessible as a companion toolbox which comprises a series of workflows to efficiently build, quantify and explore simulation input and output files. While the framework has been primarily designed for research, its simplicity of use and portability makes it a great tool for teaching purposes.

A unified framework for modelling sediment fate from source to sink and its interactions with reef systems over geological times.

Understanding the effects of climatic variability on sediment dynamics is hindered by limited ability of current models to simulate long-term evolution of sediment transfer from source to sink and associated morphological changes. We present a new approach based on a reduced-complexity model which computes over geological time: sediment transport from landmasses to coasts, reworking of marine sediments by longshore currents, and development of coral reef systems. Our framework links together the main sedimentary processes driving mixed siliciclastic-carbonate system dynamics. It offers a methodology for objective and quantitative sediment fate estimations over regional and millennial time-scales. A simulation of the Holocene evolution of the Great Barrier Reef shows: (1) how high sediment loads from catchments erosion prevented coral growth during the early transgression phase and favoured sediment gravity-flows in the deepest parts of the northern region basin floor (prior to 8 ka before present (BP)); (2) how the fine balance between climate, sea-level, and margin physiography enabled coral reefs to thrive under limited shelf sedimentation rates after ~6 ka BP; and, (3) how since 3 ka BP, with the decrease of accommodation space, reduced of vertical growth led to the lateral extension of reefs consistent with available observational data.

Guatemala paleoseismicity: from Late Classic Maya collapse to recent fault creep.

We combine 'on-fault' trench observations of slip on the Polochic fault (North America-Caribbean plate boundary) with a 1200 years-long 'near-fault' record of seismo-turbidite generation in a lake located within 2 km of the fault. The lake record indicates that, over the past 12 centuries, 10 earthquakes reaching ground-shaking intensities ≥ VI generated seismo-turbidites in the lake. Seismic activity was highly unevenly distributed over time and noticeably includes a cluster of earthquakes spread over a century at the end of the Classic Maya period. This cluster may have contributed to the piecemeal collapse of the Classic Maya civilization in this wet, mountainous southern part of the Maya realm. On-fault observations within 7 km of the lake show that soils formed between 1665 and 1813 CE were displaced by the Polochic fault during a long period of seismic quiescence, from 1450 to 1976 CE. Displacement on the Polochic fault during at least the last 480 years included a component of slip that was aseismic, or associated with very light seismicity (magnitude <5 earthquakes). Seismicity of the plate boundary is therefore either non-cyclic, or dominated by long-period cycles (>1 ky) punctuated by destructive earthquake clusters.