Search

During the Anthropocene, the impact of humans on Earth surface processes has increased exponentially, often surpassing the importance of natural drivers. Also in mountainous areas, landslide mobilization rates are exacerbated by human disturbances of the landscape such as deforestation, road constructions, and mining processes. However, investigating these interactions remains difficult in many regions due to a lack of sufficiently long observation periods, preferably over a large area, so that the presence of extreme landslide events (triggered by rainfall or earthquakes) does not induce an observation bias. Here, we investigate landslide mobilization rates in the densely populated North Tanganyika-Kivu Rift Region (NTK Rift), a prominent landslide hotspot in Africa. We use ca. 2,400 panchromatic aerial photographs from 1958 in combination with recent satellite imagery to assess the long-term landslide mobilization rates over a large area of ca. 21,000 km2.By estimating the volume of the deep-seated and shallow rapidly-moving landslides using empirical volume-area scaling relationships, we estimate that the average landslide mobilization rate in the NTK Rift is ca. 31 m3 km-2 year-1 in actively incising, rejuvenating landscapes and ca. 12 m3 km-2 year-1 in relict landscapes. The mobilization rates in the NTK Rift are dominated by the largest landslides. For instance, the 15 largest deep-seated landslides account for 50% of the total rate. Overall, we observe mobilization rates in the NTK Rift that are somewhat lower than what a global model predicts. These relatively low rates could be explained by four factors: (i) the absence of major landslide-triggering earthquakes during our 60-year observation period, (ii) the exclusion of earthflows from our analysis due to a lack of information on the depth and velocity of these instances, (iii) the relatively large size of our study area which reduces biases linked to extreme rainfall, (iv) the fact that the NTK Rift is a mountain range in an extension area, which differs from orogenic mountainous areas, where most landslide mobilization rates are reported; and (v) uncertainties on the global landslide mobilization rate model.In rejuvenated landscapes, roughly 5% of the sediment mobilization by rapidly-moving landslides is linked to human activity, while in relict landscapes this figure rises to 18%, notably due to mining and road construction. The role of human activity is limited as compared to the recent occurrence of some large landslides, which seem linked to natural causes and dominate the overall mobilization rates. Moreover, the limited role of human activity must be balanced with the fact that the NTK Rift, although highly populated, remains relatively untouched by major road infrastructure constructions. While previous studies have found that deforestation has a large impact on the landside risk (i.e. the incidence of landslide fatalities), its impact on the observed mobilization rates appears to be much less important. The landslides associated with deforestation are commonly shallow debris avalanches with a limited size and rather high mobility.Overall, our results significantly contribute to a better understanding of landslide mobilization and its controlling factors, especially by proving much-needed long-term observations for a currently under-researched type of environment.

Tropical mountainous regions are often identified as landslide hotspots with particularly vulnerable populations. Anthropogenic factors are assumed to play a role in the occurrence of landslides in these populated regions, yet the relative importance of these human-induced factors remains poorly documented. In this work, we aim to explore the impact of forest cover dynamics, roads and mining activities on the occurrence of landslides in the Rift flank west of Lake Kivu in the DR Congo. To do so, we compile an inventory of 2730 landslides using © Google Earth imagery, high resolution topographic data, historical aerial photographs from the 1950’s and extensive field surveys. We identify old and recent (post 1950’s) landslides, making a distinction between deep-seated and shallow landslides, road landslides and mining landslides. We find that susceptibility patterns and area distributions are different between old and recent deep-seated landslides, which shows that natural factors contributing to their occurrence were either different or changed over time. Observed shallow landslides are recent processes that all occurred in the past two decades. The analysis of their susceptibility indicates that forest dynamics and the presence of roads play a key role in their regional distribution pattern. Under similar topographic conditions, shallow landslides are more frequent, but of smaller size, in areas where deforestation has occurred since the 1950’s as compared to shallow landslides in forest areas, i.e. in natural environments. We attribute this size reduction to the decrease of regolith cohesion due to forest loss, which allows for a smaller minimum critical area for landsliding. In areas that were already deforested in 1950’s, shallow landslides are less frequent, larger, and occur on less steep slopes. This suggests a combined role between regolith availability and soil management practices that influence erosion and water infiltration. Mining activities increase the odds of landsliding. Mining and road landslides are larger than shallow landslides but smaller than the recent deep-seated instabilities. The susceptibility models calibrated for shallow and deep-seated landslides do not predict them well, highlighting that they are controlled by environmental factors that are not present under natural conditions. Our analysis demonstrates the role of human activities on the occurrence of landslides in the Lake Kivu region. Overall, it highlights the need to consider this context when studying hillslope instability characteristics and distribution patterns in regions under anthropogenic pressure. Our work also highlights the importance of considering the timing of landslides over a multi-decadal period of observation.

In the version of this Article initially published, an error appeared in the caption to Fig. 4, where the descriptions for panels b and c were reversed. The caption has been corrected to read: “b, Landslide hazard trends when we assume there is no landslide response to deforestation in new forests. c, Landslide hazard trends in the Kivu Rift when we assume the landslide response to deforestation is the same for new and primary forests.” The error has been corrected in the online version of the Article.

Deforestation is associated with a decrease in slope stability through the alteration of hydrological and geotechnical conditions. As such, deforestation increases landslide activity over short, decadal timescales. However, over longer timescales (0.1–10 Myr) the location and timing of landsliding is controlled by the interaction between uplift and fluvial incision. Yet, the interaction between (human-induced) deforestation and landscape evolution has hitherto not been explicitly considered. We address this issue in the North Tanganyika–Kivu rift region (East African Rift). In recent decades, the regional population has grown exponentially, and the associated expansion of cultivated and urban land has resulted in widespread deforestation. In the past 11 Myr, active continental rifting and tectonic processes have forged two parallel mountainous rift shoulders that are continuously rejuvenated (i.e., actively incised) through knickpoint retreat, enforcing topographic steepening. In order to link deforestation and rejuvenation to landslide erosion, we compiled an inventory of nearly 8000 recent shallow landslides in © Google Earth imagery from 2000–2019. To accurately calculate landslide erosion rates, we developed a new methodology to remediate inventory biases linked to the spatial and temporal inconsistency of this satellite imagery. Moreover, to account for the impact of rock strength on both landslide occurrence and knickpoint retreat, we limit our analysis to rock types with threshold angles of 24–28∘. Rejuvenated landscapes were defined as the areas draining towards Lake Kivu or Lake Tanganyika and downstream of retreating knickpoints. We find that shallow landslide erosion rates in these rejuvenated landscapes are roughly 40 % higher than in the surrounding relict landscapes. In contrast, we find that slope exerts a stronger control on landslide erosion in relict landscapes. These two results are reconciled by the observation that landslide erosion generally increases with slope gradient and that the relief is on average steeper in rejuvenated landscapes. The weaker effect of slope steepness on landslide erosion rates in the rejuvenated landscapes could be the result of three factors: the absence of earthquake-induced landslide events in our landslide inventory, a thinner regolith mantle, and a drier climate. More frequent extreme rainfall events in the relict landscapes, and the presence of a thicker regolith, may explain a stronger landslide response to deforestation compared to rejuvenated landscapes. Overall, deforestation initiates a landslide peak that lasts approximately 15 years and increases landslide erosion by a factor 2 to 8. Eventually, landslide erosion in deforested land falls back to a level similar to that observed under forest conditions, most likely due to the depletion of the most unstable regolith. Landslides are not only more abundant in rejuvenated landscapes but are also smaller in size, which may again be a consequence of a thinner regolith mantle and/or seismic activity that fractures the bedrock and reduces the minimal critical area for slope failure. With this paper, we highlight the importance of considering the geomorphological context when studying the impact of recent land use changes on landslide activity.

Tropical mountainous regions are often identified as landslide hotspots with particularly vulnerable populations. Whereas both natural (e.g., rainfall, lithology) and anthropogenic (e.g., deforestation, mining) factors are expected to be involved in landslide occurrence, the relative importance of such factors remains poorly documented. The objective of this study is to understand the role played by natural and anthropogenic factors in the occurrence of landslides in the Rift Flank west of Lake Kivu in DR Congo. First, we inventoried 2856 landslides of various types using Google Earth imagery, high resolution topographic data, historical aerial photographs from1950’sand intense field surveys. We made a distinction between old and recent deep-seated landslides, shallow landslides and mining-related landslides. The difference in susceptibility patterns and in size distributions between old and recent deep-seated landslides indicates that natural factors contributing to their occurrence were either different or changed over time. The shallow landslides are all recent. The susceptibility analysis shows that their regional pattern is mainly controlled by forest dynamics and the presence of roads. The occurrence of shallow landslides in forest areas, i.e. in natural environments, strongly increases with slope gradient. In areas with similar topographic conditions where deforestation has occurred since the 1950’s, shallow landslides are more frequent, but of smaller size. This size reduction is attributed to the decrease of regolith cohesion due to forest loss, which allows a smaller minimum critical area for landsliding. In areas that were already deforested in 1950’s, shallow landslides are less frequent, larger, and occur on less steep slopes. This suggests a combined role between regolith availability and soil management practices that influence erosion and water infiltration. Landslides associated with mining activities are larger than shallow landslides but smaller than the recent deep-seated instabilities. They are not well predicted by the susceptibility models, showing that they respond to environmental factors that are not present under natural conditions. This research highlights the importance of human activities on the occurrence of landslides and the need to consider this context when studying hillslope instability patterns in regions under anthropic pressure. Keywords: landslide processes, multi-temporal inventory, historical deforestation, mining and quarrying, susceptibility assessment, Africa

The understanding of the interplay between natural and human induced factors in the occurrence of landslides remains poorly constrained in many regions, especially in tropical Africa where data-scarcity is high. In these regions where population growth is significant and causes changes in land use/cover, the need for a sustainable management of the land is on the rise. Here, we aim to unravel the occurrence of landslides in the 40 km-long Ruzizi gorge, a rapidly incising bedrock river in the Kivu Rift in Africa that has seen its landscape disturbed over the last decades by the development of the city of Bukavu (DR Congo). Careful field observations, historical aerial photographs, satellite imagery and archive analysis are combined to produce a multi-temporal inventory of 264 landslides. We show that the lithological context of the gorge and its extremely high incision rate (> 20 mm year-1) during the Holocene explains the presence of a concentration of large landslides (up to 2 km²) of undetermined age (well before the first observations of 1959) whose occurrence is purely natural. They are mostly of the slide type and do not show morphologic patterns of recent activity. The landslides that occurred during the last 60 years are flow-like shallower slope failures of smaller size (up to 0.12 km²) and tend to disappear rather quickly (sometimes within a few years) from the landscape as a result of rapid vegetation growth, land reclamation and (human-induced) soil erosion. They are primarily related to threshold slopes and precipitation plays a frequent role in their onset. However, land use/cover changes also affect their occurrence. This study provides useful information for a more accurate evaluation of the landslide hazard in the area, particularly with respect to the growth of the city of Bukavu that has developed without the consideration of naturally instable slopes. It also stresses the need and added value of building accurate landslide inventories in data-scarce regions.

On the nexus of humans and their environment, landslide risk is in essence dynamic. In mountainous areas over the world, the need for agricultural land incites people to settle on steeper (more landslide-prone) terrain at the expense of ecosystems. At the same time, the degradation of ecosystems, for example through deforestation, leads to a considerable increase in landslide hazard. Although the link between deforestation and landslide hazard/risk has been widely recognized, it remains poorly quantified. This is especially the case in the Global South where historical land cover and landslide records are scarce. In this study, we investigate 58 years of forest cover changes, population dynamics, and landslide risk in the Kivu Rift. This mountainous region presents similar geomorphic and climatic conditions across three countries: Burundi, the eastern part of the Democratic Republic of the Congo (DRC), and Rwanda. First, we use contemporary landslide and deforestation data (2000-2016) to explicitly quantify the interactions between these two processes. Second, we reconstruct the annual forest cover changes between 1958 and 2016 by means of a cellular automaton of which the output converges to four forest cover products (1958, 1988, 2001, 2016). We derive the 1958 forest data from an inventory of nearly 2,400 panchromatic aerial photographs, available at the Royal Museum for Central Africa. The forest data for 1988, 2001, and 2016 are readily available and derived from satellite imagery. Next, we estimate the yearly historical landslide hazard dynamics by applying the contemporary deforestation-landslide relationship to the historical forest cover changes. Finally, an approximation of the landslide risk (expected fatalities per 100,000 inhabitants), is calculated for four epochs (1975, 1990, 2000, 2015) and derived from the product of the corresponding hazard map and population density grids.During our entire period of observation, the landslide risk is higher in the DRC than in Rwanda and Burundi. While the risk in Rwanda and Burundi displays a slightly decreasing trend, the risk seems more volatile in the DRC. Here, the initial risk in 1975 is high due to the concentration of a small population along the steep northwestern coast of Lake Kivu. In the following 15 years, the risk in the DRC decreases sharply, only to soar again in the nineties. This sudden increase in risk can be linked to two factors: demographic changes and environmental degradation. During the nineties, the location of the Congolese people shifted towards steeper terrain. This shift is explained by the relocation of hundreds of thousands of Rwandan refugees and internally displaced people following the First and Second Congo War, but also by the economic opportunities provided by the booming, often informal, mining industry. Deforestation has also contributed to the higher landslide risk in the DRC, as large parts of the primary forest have been cut to satisfy the land and fuelwood demand of the fast-growing population.With our analysis, we demonstrate that a landslide risk assessment is more than the reflection of the current environmental conditions. The legacy of environmental and societal dynamics resonates in contemporary landslide risk.

Deforestation increases landslide activity over short, contemporary timescales. However, over longer timescales the location and timing of landsliding is controlled by the interaction between uplift and fluvial incision. Yet, the interaction between (human-induced) deforestation and landscape evolution has hitherto not been explicitly considered. We address this issue in the North Tanganyika-Kivu Rift region (East African Rift). In recent decades, the regional population has grown exponentially and the associated expansion of cultivated and urban land has resulted in widespread deforestation. On a much longer time scale, tectonic uplift has forged two parallel mountainous Rift shoulders that are continuously rejuvenated through knickpoint retreat, enforcing topographic steepening. In order to link deforestation and rejuvenation to landslide erosion, we compiled an inventory of nearly 8,000 recent shallow landslides in Google Earth© imagery from 2000–2019. To accurately calculate landslide erosion rates, we developed a new methodology to remediate inventory biases linked to the spatial and temporal inconsistency of this satellite imagery. We find that erosion rates in rejuvenated landscapes are roughly 40 % higher than in the surrounding relict landscapes, upstream of retreating knickpoints and outside of the Rift shoulders. This difference is due to the generally steeper relief in rejuvenated landscapes which more than compensates for the fact that rejuvenated slopes, when compared to similarly angled slopes in relict zones, often display a somewhat lower landslide erosion rate. These lower rates in the rejuvenated landscapes could be the result of a drier climate, the omission of earthquake-induced landslide events in our landslide inventory, and potentially a smaller regolith stock. More frequent extreme rainfall events in the relict zones, and possibly the presence of a thicker regolith, cause a stronger landslide response to deforestation compared to rejuvenated landscapes. Overall, deforestation initiates a landslide peak that lasts approximately 15 years and increases landslide erosion by a factor 2 to 8. Eventually, landslide erosion in deforested land falls back to a level similar to that observed under forest conditions, most likely due to the depletion of the most unstable regolith. Landslides are not only more abundant in rejuvenated landscapes but are also smaller in size, which may be a consequence of the seismic activity that fractures the bedrock and reduces the minimal critical area for slope failure. With this paper, we highlight the importance of considering the geomorphological context when studying the impact of recent land use changes on landslide activity.

Both landscape rejuvenation through tectonic uplift and human-induced deforestation are known to increase landslide (LS) activity. Yet, the interaction between deforestation and landscape evolution has hitherto not been explicitly considered. Here, we investigate how shallow LS frequency is impacted by deforestation and landscape rejuvenation through knickpoint retreat in the Kivu Rift (East African Rift) while accounting for rock strength and slope steepness. In the past 12 Ma, the Kivu Rift has been characterized by tectonic uplift which gave rise to knickpoints in the river profiles enforcing topographic steepening. On a much shorter timescale, the rapidly growing population in the Rift has gradually expanded its cultivated and urban land leading to widespread deforestation.We compiled an inventory of almost 8000 shallow LSs using Google Earth imagery. To quantify LS frequency, we developed a new method that accounts for the temporal and spatial inconsistency of satellite imagery coverage. To characterize long-term landscape evolution, we identified (i) 672 non-stationary knickpoints in the Rift and (ii) quantified the impact of lithology on slope threshold angles (TA). We identified two homogenous lithological groups: one group of younger/weaker lithologies (540 Ma, TA=27.9 +/- 0.3°). Further analysis focused on the latter group since it covers 85% of the study area and contained more than 95% of the observed LSs.The overall shallow LS frequency in the rejuvenated landscapes inside the rift is 0.039 LS/km2/yr versus 0.010 LS/km2/yr in the relict landscapes outside the rift. Generally, LS frequency on recently deforested slopes increased by 200 to 800% in comparison to forested land. There is no notable difference in LS frequency on equally steep non-forested slopes (i.e. slopes deforested at least several decades ago) inside and outside of the rift. However, forest slopes of similar steepness are 2-3 times more sensitive to landsliding within the rift. We propose two mechanisms that might explain the higher frequency of landsliding on similar topographies within the rift: (i) the active undercutting by rivers may lead to slope destabilization without significantly increasing the average slope gradient as extracted from the SRTM DEM and (ii) tectonic uplift may induce rock and regolith fracturing, leading to weaker, more LS-prone slopes. The fact that we did not observe differences in LS frequency on hillslopes that were deforested long ago suggests that on such slopes a new equilibrium is established whereby these aforementioned mechanisms are no longer important.In conclusion, one of the key factors why the rejuvenated landscape inside the rift is more sensitive to landsliding is the higher prevalence of threshold slopes due to active incision. However, the impact of rejuvenation cannot be understood by considering only its effects on overall topography. Deforestation dramatically increases LS frequency in both relict and rejuvenated landscapes, in the first decades after forest cover removal.