Report

Latitudinal effect of vegetation on erosion rates identified along western South America

See allHide authors and affiliations

Science  20 Mar 2020:
Vol. 367, Issue 6484, pp. 1358-1361
DOI: 10.1126/science.aaz0840

Erosion-vegetation interactions

The impact of vegetation on erosion rates is hard to gauge. Although vegetation can hold soils in place mechanically, root systems can also loosen soils or even help to fracture rock. These processes can increase erosion, especially because areas of heavy vegetation tend to be in areas with high precipitation rates. Starke et al. tackled this issue using a large set of observations that span 3500 km of the Andes mountain range. They found a complex set of interactions where increasing vegetation decreases erosion in more arid regions but can accelerate erosion in vegetation dense regions.

Science, this issue p. 1358

Abstract

Vegetation influences erosion by stabilizing hillslopes and accelerating weathering, thereby providing a link between the biosphere and Earth’s surface. Previous studies investigating vegetation effects on erosion have proved challenging owing to poorly understood interactions between vegetation and other factors, such as precipitation and surface processes. We address these complexities along 3500 kilometers of the extreme climate and vegetation gradient of the Andean Western Cordillera (6°S to 36°S latitude) using 86 cosmogenic radionuclide–derived, millennial time scale erosion rates and multivariate statistics. We identify a bidirectional response to vegetation’s influence on erosion whereby correlations between vegetation cover and erosion range from negative (dry, sparsely vegetated settings) to positive (wetter, more vegetated settings). These observations result from competing interactions between precipitation and vegetation on erosion in each setting.

The impact of vegetation on the shape and evolution of Earth’s surface ranges (for example) from the microscopic scale of Mycorrhiza weathering for plant nutrition to macroscopic scales where plants retard hillslope erosion, stabilize environments for sediment deposition, and affect precipitation through evapotranspiration, interception, and leaf phenology (16). However, defining the influence of vegetation on catchment-averaged erosion rates has proven difficult because of, among other things, nonlinear interactions between vegetation type and cover with precipitation, temperature, and solar radiation (710). One approach for disentangling the effects of vegetation and climate on landscape evolution requires quantifying catchment erosion rates over a large range of climate and biogeographic conditions. The production of cosmogenic radionuclides in the upper ~2 m of Earth’s surface provides one means for quantifying millennial time scale catchment-averaged erosion rates (11, 12). We investigate the relationships between catchment-averaged erosion rates with vegetation cover, climate, and topographic slope along the climate and ecological gradient of the Andean Western Cordillera, South America.

The Andean Western Cordillera between 6°S and 36°S latitude extends 3500 km (Fig. 1A) and crosses six climate zones, from hyperarid to temperate (13), and four distinct biogeographic regions (Fig. 1C and 2A) (14). Cosmogenic radionuclide–derived erosion rates and their controlling factors have been investigated along the Andean Western Cordillera, often with conflicting results [e.g., (15, 16)]. The emphasis of previous studies ranged from quantification of erosion rates in the vegetation-limited Atacama Desert (16, 17) and sediment storage in hyperarid environments (18) to the rates of canyon incision and hillslope erosion in the Andean Western Cordillera (19). Few studies (15, 20) have previously looked at systematic latitudinal variations in erosion rates along this gradient.

Fig. 1 Topography, erosion rates, and vegetation types along the western Andean margin.

(A) Topographic map showing the catchment-averaged erosion rate sample locations of river sediments from the Andean Western Cordillera. Black dots indicate published 10Be concentrations. Red dots are new data presented in this study (tables S1 and S2). (B) Calculated catchment-averaged erosion rates (m/Myr) with 1σ uncertainty versus latitude (°S). The black line represents the three-point moving average. All catchment-averaged erosion rates were calculated with the same procedure (materials and methods). (C) Percent vegetation type versus latitude across a 100-km-wide latitudinal profile in the Andean Western Cordillera. Gray lines represent barren or sparsely vegetated areas. Black lines denote shrublands. Green lines represent grasslands, and purple lines indicate woody savannas. Blue lines represent mixed forests and orange lines, evergreen forest. Values are derived from MODIS 2012 vegetation continuous field data.

Fig. 2 Latitudinal variations in vegetation cover, precipitation, slope, and their correlation with erosion rates from the catchments shown in Fig. 1A.

(A) Vegetation cover and type plotted versus latitude. Vegetation types shown in colored zones are from MODIS 2012 vegetation continuous field data (see also Fig. 1C). The vegetation types represent mixed forest (I), grassland (II), open shrubland (III), and barren or sparsely vegetated areas (IV). (B) TRMM2b precipitation versus latitude. (C) Mean catchment slope (90 m window) versus latitude. All dashed lines in (A) to (C) represent the three-point moving averages of the 2σ variation from the mean of each value. Solid lines in (A) to (C) represent the three-point moving averages of values. (D) Correlation coefficients (R) versus latitude for erosion rate and slope (red) and erosion rate and precipitation (blue). Dots and color-shaded regions show the mean and 1σ uncertainty within each bin based on a Monte Carlo analysis of the variability in erosion rates and precipitation or slope within each 2° bin (see tables S3 to S5 for all values plotted). (E) Correlation coefficient versus latitude for erosion rates and vegetation cover. Dots and color-shaded regions show mean and 1σ uncertainty, as in (D.) The dashed green line represents trends in the correlation coefficient with latitude.

We measured cosmogenic radionuclide concentrations of 10Be (12) from 12 samples and combined this data with 74 published samples from Peru and Chile (Fig. 1, A and B, and tables S1 and S2). The 86 catchments are adjacent to a similar tectonic plate boundary with 13 (broadly defined) catchment lithologies, including Oligo-Miocene, Plio-Pleistocene volcanoclastic deposits and ignimbrites, Jurassic and Cretaceous sedimentary rocks, Paleozoic and Cretaceous granodiorites, and Precambrian gneiss (21). The total lithological-weighted quartz content for each catchment varies between 15 to 49% (fig. S2). We used 10Be concentrations to recalculate erosion rates using the same sea level high-latitude production rates and production-rate scaling (22) (fig. S1A). We determined from MODIS, TRMM, CHELSA, WolrdClim, SRTM, and GLiM datasets (Figs. 1C and 2 and fig. S3) the 2σ range in vegetation cover, mean annual precipitation (MAP) and temperature (MAT), mean solar radiation (MSR), catchment-averaged slope, local relief, and lithologic quartz content for each catchment. We found no large differences between the TRMM, CHELSA, and WorldClim (MAT and MAP values) datasets (23) (tables S1 to S6).

The catchment-averaged erosion rates varied between 1.4 and 150 m per million years (m/Myr) (solid line, Fig. 1B). Starting in the north (6°S to 12°S), erosion rates displayed increasing values between 0 and 150 m/Myr. From 12°S to 20°S, the erosion rates decreased (150 to 0 m/Myr). The lowest erosion rates were located between 20°S to 30°S (0 to 50 m/Myr). In the south (30°S to 36°S), erosion rates ranged between 0 to 140 m/Myr and showed increasing values from 30°S to 33.5°S and decreasing values from 33.5°S to 36°S. In general, the quartz content for each catchment shows no latitudinally dependent variation (fig. S2C). Vegetation cover and MAP were the highest in the north from 6°S to 10°S (50 to 85% and 200 to 700 mm/year, respectively; solid lines in Fig. 2, A and B). Vegetation cover and MAP then decreased to a minimum (5% and <50 mm/year) at the latitudes of the Atacama Desert (20°S to 30°S). Further south (30°S to 36°S), the vegetation cover and MAP increased to a southern maximum (40% and ~700 mm/year). Catchment-averaged slopes had increasing values up to 30° from 6°S to 12°S. Slopes gradually decreased toward the south (12°S to 20°S) and varied between 25° to 10°. The lowest slopes (5° to 10°) were situated between 20°S to 30°S and increased to the south (30°S to 36°S) up to 28° (solid line, Fig. 2C).

Results from a multivariate factor analysis indicated that catchment erosion rate, vegetation cover, slope, MAT, MAP, MSR, local relief, quartz content, and lithology result in four factors that can explain 62% of the variance (tables S4 and S5). Factor 1 explains 23% of the variance and contains MAP, vegetation cover, erosion rate (factor loadings of 0.9, 0.6, and 0.5, respectively). We interpreted factor 1 as representing the combined interactions of vegetation, precipitation, and erosion. Factor 2 (20% of the variance) contains vegetation cover, MAT (factor loading >0.6), and acid volcanic rocks (“va” in table S4, factor loading 0.7). We interpreted factor 2 as representing interactions between vegetation, temperature, and substrate composition. Both the factor and Pearson statistical analyses showed no correlation or covariance of erosion rates with quartz content or rock type (tables S4 to S6 and figs. S1 and S6).

We report our results for the Pearson correlation coefficients (R) averaged over n catchments within 2° latitudinal increments (Fig. 2, D and E). The Pearson R values are a metric for the degree of linear dependence between individual parameters and the erosion rate (e.g., fig. S9). We chose the 2° increments as the minimum spatial scale over which a sufficient number of catchments (n > 5) are available for analysis (e.g., fig. S9 and table S5). The correlation coefficients are reported for a range of values in each bin using a Monte Carlo analysis of the 2σ range of uncertainties in a catchment (Figs. 1 and 2). We classified the absolute values of the correlation coefficients into regions with very weak (0.00 to 0.19), weak (0.20 to 0.39), moderate (0.40 to 0.59), strong (0.60 to 0.79), and very strong (0.80 to 1.0) R values (23).

We found that precipitation had a clear latitudinal gradient, but the correlation coefficients between precipitation and erosion rate do not follow this gradient. Instead, they oscillate between very weak to moderate (Fig. 2D). Similarly, no clear systematic latitudinal variation in the correlation coefficients existed between slope and erosion rates (except between 6°S and 12°S). By contrast, vegetation-erosion correlations show a pattern of latitudinal variations to the north and south of the arid region (region A, Fig. 2E, 18°S to 32°S). More specifically, in the arid (<100 mm/year) and sparsely vegetated region A (vegetation cover <20%), the vegetation-erosion correlation indicated a very weak to moderate negative relationship. To the north and south of region A, the correlation between vegetation and erosion rate was near 0 and very weak to moderate (region B1, Fig. 2E). Further north (region C, 14°S to 10°S), in the more vegetated (50 to 60% vegetation cover) and wetter (200 to 300 mm/year precipitation) latitudes, a positive strong to very strong correlation was found. We could not document a similar maximum in the vegetation-erosion correlation (or region C) in the south from 36°S to 40°S because we lacked samples from the region. In the most heavily vegetated (50 to 80% cover) and wettest (300 to 700 mm/year precipitation) area north of 10°S (region B2), the correlation strength is very weak and similar to that in region B1 (Fig. 2E).

We interpreted latitudinal changes in the vegetation-erosion correlation strength (regions A to C, Fig. 2E) qualitatively on the basis of abiotic and biotic factors that influence catchment erosion. Region A (18°S to 32°S) has sparse vegetation cover (<20%) and arid conditions. Stochastic variations in precipitation occur in this region over decadal time scales such that we observed, with one exception, a positive correlation with erosion rates and precipitation (Fig. 2D). A moderate negative vegetation-erosion correlation occurred in region A, which indicated that the sparse vegetation present was sufficient to influence erosion rates by increasing surface roughness for overland flow (Fig. 2E). Region B1 (14°S to 18°S and 32°S to 36°S) has a vegetation cover of 20 to ~50% and represents a transition zone where a very weak to weak erosion-vegetation correlation occurs. This weak vegetation-erosion correlation could be due to competing processes such that increased biotic regolith production associated with higher vegetation cover is offset by the ability of vegetation cover to retard the physical transport of sediment. Alternatively, the weak correlation between erosion rates and vegetation in region B1 (and also B2) could reflect areas where landscapes were closer to steady state with the rock uplift rate. In steady-state landscapes, correlations between erosion rate and vegetation, or precipitation, are expected to be weak. In this case, erosion rates should be positively correlated with slope. This interpretation is partially supported in regions B1 and B2. In region C (14°S to 10°S, Fig. 2E), the observed maximum in the erosion rate and vegetation correlation indicated that vegetation contributed to enhanced weathering and erosion but was ineffective at stabilizing hillslopes from erosion under the higher (~300 mm/year) precipitation rates. Finally, the decrease in the correlation to near zero in the northernmost area (region B2) indicated that dense vegetation cover of 50 to 85% hindered erosion such that the correlation coefficient between vegetation and erosion decreased (24). We also found that vegetation cover greater than 50% leads to a maintenance of steep mean slopes (25° to 30°; fig. S6). By contrast, regions A and B1 had lower mean slope angles between 5° and 25°.

Our observations are consistent with coupled vegetation-landscape evolution modeling work that investigated the effects of varying precipitation and vegetation cover on catchment erosion in Chile (25). More specifically, Schmid et al. [see figure 17 in (25)] applied a nonlinear parameterization for vegetation cover–dependent fluvial erodibility and hillslope diffusivity. They found an inverse relationship (negative correlation) between catchment erosion and vegetation cover for sparsely vegetated areas (~10% vegetation cover) due to small increases in precipitation in an arid setting resulting in an increase in vegetation cover that reduced erosion. The negative vegetation-erosion correlation and positive precipitation-erosion correlation that we observed in region A (Fig. 2, D and E) are comparable to this modeling result (25). By contrast, in more vegetated regions (~70% vegetation cover), the model results suggest that precipitation and vegetation cover are positively correlated with catchment erosion. This trend is caused by the high precipitation rates required to sustain dense vegetation cover having a stronger impact on runoff-related erosion and biotic weathering than the stabilizing effects of vegetation cover on erosion. This prediction is consistent with the positive vegetation-erosion and precipitation-erosion correlations that we observed in regions C (~10°S to 14°S, 50 to 70% vegetation cover, Fig. 2E). The northward decrease in the vegetation-erosion correlation (region B2) that we observed reflects that above ~50 to 70% vegetation cover, vegetation effects on inhibiting erosion outpace increases in the precipitation rates that promote erosion. Thus, both the observations presented here and landscape evolution model results (25) confirm a bidirectional response of vegetation cover and precipitation effects on catchment erosion. Similar findings have been reported for smaller geographic areas in both East Africa (26) and the Himalaya (27), as well as globally for differing amounts of tree cover (20).

Complications associated with our interpretations could result from latitudinal variations in tectonic activity, paleoclimate, and paleovegetation cover. However, the regions of vegetation-erosion interactions that we identified (Fig. 2) do not correspond to known patterns of upper plate seismicity or subducting oceanic ridges (28) (fig. S7). The main phases of mountain building in the Andean Western Cordillera were from 20 to 10 million years ago (Ma) and terminated around 9 Ma (29, 30), near the time when river knick-points initiated and migrated upstream (31) (fig. S8). Another potential caveat is that paleoprecipitation rates and paleovegetation cover could differ in magnitude from the modern values. However, paleo-precipitation gradients (from Pliocene to modern time) in the region are similar to modern precipitation gradients along the Andean Western Cordillera (32) (fig. S5), and latitudinal variations in paleovegetation cover in north-central Chile are estimated to be within 5 to 15% of the present-day values (33). Finally, this study focuses on regional-scale variations in vegetation cover and catchment erosion. This simplified approach highlights the need to evaluate how individual plant functional types (e.g., trees, shrubs, grasses), and not just total vegetation cover, could affect biotic weathering and erosional processes [e.g., (2, 20, 33)].

Our observations have broader implications for the vegetation cover effects on erosion. The latitudinal variations that we identified in the correlation strength between vegetation, precipitation, and erosion imply that studies conducted on a smaller spatial scale could poorly resolve vegetation-erosion interactions. For example, Fig. 3 shows that over the entire study area, the magnitude, and variance, of erosion rates observed decrease with increased vegetation cover. If we take a subset of samples from the catchments shown in Fig. 3, then recovering correlations between vegetation and erosion would be difficult, particularly in regions with low (≲60%) vegetation cover, where the variance in erosion rates is high. This conclusion helps explain the diversity of vegetation-erosion relationships synthesized in previous work (34, 35). Results from previous studies (15, 26, 36) have shown both positive and negative correlations between vegetation cover, precipitation, and erosion rates. These seemingly conflicting results may have occurred in areas that are located at, or straddle, the diverse range of climate and vegetation cover regions identified here (Figs. 2E and 3).

Fig. 3 Observed relationship between erosion rate, vegetation cover, and mean annual precipitation rate (colors) for each catchment in Figs. 1 and 2.

Black squares represent outliers that are possibly biased by glaciation in the catchment. For comparison, fig. S6A is identical but color-coded by slope. Regions labeled correspond to those identified in Fig. 2E.

In conclusion, we found that correlations between catchment erosion and vegetation cover spanning 30° latitude varied in their direction (positive or negative) and strength (very weak to very strong). These observations, taken together with previous modeling results, indicated a bidirectional relationship between vegetation cover and catchment erosion. The source of this nonlinearity is due to competing, and latitudinally varying, interactions between precipitation and vegetation on erosion. Results presented here provide a regional context for future (smaller scale) studies investigating similar interactions over a narrower range of vegetation cover and precipitation.

Supplementary Material

science.sciencemag.org/content/367/6484/1358/suppl/DC1

Materials and Methods

Figs. S1 to S9

Tables S1 to S6

Google Earth kmz file of catchment properties

References (3768)

References and Notes

Acknowledgments: We thank D. Kost and L. Michel for laboratory and field assistance, respectively. We also thank three anonymous reviewers and R. Drews for their thoughtful comments. Funding: This study was funded by a European Research Council (ERC) consolidator grant (CoG 615703) and the German Science Foundation (DFG) priority research program EarthShape (EH329/17-2) to T.A.E. Author contributions: T.A.E. and J.S. planned the study. J.S. was responsible for all sample collection and calculations. J.S. and M.S. performed the laboratory analysis. All authors contributed to manuscript and figure preparation. Competing interests: The authors have no competing financial conflicts of interest with this study. Data and materials availability: Data reported in the paper are presented in the supplementary materials.

Stay Connected to Science

Navigate This Article