@article{ZimmermannFranckeElsenbeer2012, author = {Zimmermann, Alexander and Francke, Till and Elsenbeer, Helmut}, title = {Forests and erosion: Insights from a study of suspended-sediment dynamics in an overland flow-prone rainforest catchment}, series = {Journal of hydrology}, volume = {428}, journal = {Journal of hydrology}, number = {7}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0022-1694}, doi = {10.1016/j.jhydrol.2012.01.039}, pages = {170 -- 181}, year = {2012}, abstract = {Forests seem to represent low-erosion systems, according to most, but not all, studies of suspended-sediment yield. We surmised that this impression reflects an accidental bias in the selection of monitoring sites towards those with prevailing vertical hydrological flowpaths, rather than a tight causal link between vegetation cover and erosion alone. To evaluate this conjecture, we monitored, over a 2-year period, a 3.3 ha old-growth rainforest catchment prone to frequent and widespread overland flow. We sampled stream flow at two and overland flow at three sites in a nested arrangement on a within-event basis, and monitored the spatial and temporal frequency of overland flow. Suspended-sediment concentrations were modeled with Random Forest and Quantile Regression Forest to be able to estimate the annual yields for the 2 years, which amounted to 1 t ha(-1) and 2 t ha(-1) in a year with below-average and with average precipitation, respectively. These estimates place our monitoring site near the high end of reported suspended-sediment yields and lend credence to the notion that low yields reflect primarily the dominance of vertical flowpaths and not necessarily and exclusively the kind of vegetative cover. Undisturbed forest and surface erosion are certainly no contradiction in terms even in the absence of mass movements.}, language = {en} } @article{VorpahlElsenbeerMaerkeretal.2012, author = {Vorpahl, Peter and Elsenbeer, Helmut and M{\"a}rker, Michael and Schr{\"o}der-Esselbach, Boris}, title = {How can statistical models help to determine driving factors of landslides?}, series = {Ecological modelling : international journal on ecological modelling and engineering and systems ecolog}, volume = {239}, journal = {Ecological modelling : international journal on ecological modelling and engineering and systems ecolog}, number = {7}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0304-3800}, doi = {10.1016/j.ecolmodel.2011.12.007}, pages = {27 -- 39}, year = {2012}, abstract = {Landslides are a hazard for humans and artificial structures. From an ecological point of view, they represent an important ecosystem disturbance, especially in tropical montane forests. Here, shallow translational landslides are a frequent natural phenomenon and one local determinant of high levels of biodiversity. In this paper, we apply weighted ensembles of advanced phenomenological models from statistics and machine learning to analyze the driving factors of natural landslides in a tropical montane forest in South Ecuador. We exclusively interpret terrain attributes, derived from a digital elevation model, as proxies to several driving factors of landslides and use them as predictors in our models which are trained on a set of five historical landslide inventories. We check the model generality by transferring them in time and use three common performance criteria (i.e. AUC, explained deviance and slope of model calibration curve) to, on the one hand, compare several state-of-the-art model approaches and on the other hand, to create weighted model ensembles. Our results suggest that it is important to consider more than one single performance criterion. Approaching our main question, we compare responses of weighted model ensembles that were trained on distinct functional units of landslides (i.e. initiation, transport and deposition zones). This way, we are able to show that it is quite possible to deduce driving factors of landslides, if the consistency between the training data and the processes is maintained. Opening the 'black box' of statistical models by interpreting univariate model response curves and relative importance of single predictors regarding their plausibility, we provide a means to verify this consistency. With the exception of classification tree analysis, all techniques performed comparably well in our case study while being outperformed by weighted model ensembles. Univariate response curves of models trained on distinct functional units of landslides exposed different shapes following our expectations. Our results indicate the occurrence of landslides to be mainly controlled by factors related to the general position along a slope (i.e. ridge, open slope or valley) while landslide initiation seems to be favored by small scale convexities on otherwise plain open slopes.}, language = {en} } @article{GermerZimmermannNeilletal.2012, author = {Germer, Sonja and Zimmermann, Alexander and Neill, Christopher and Krusche, Alex V. and Elsenbeer, Helmut}, title = {Disproportionate single-species contribution to canopy-soil nutrient flux in an Amazonian rainforest}, series = {Forest ecology and management}, volume = {267}, journal = {Forest ecology and management}, number = {2}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0378-1127}, doi = {10.1016/j.foreco.2011.11.041}, pages = {40 -- 49}, year = {2012}, abstract = {Rainfall, throughfall and stemflow were monitored on an event basis in an undisturbed open tropical rainforest with a large number of palm trees located in the southwestern Amazon basin of Brazil. Stemflow samples were collected from 24 trees with a diameter at breast height (DBH) > 5 cm, as well as eight young and four full-grown babassu palms (Attalea speciosa Mart.) for 5 weeks during the peak of the wet season. We calculated rainfall, throughfall and stemflow concentrations and fluxes of Na+, K+, Ca2+, Mg2+,, Cl-, SO42-, NO3- and H+ and stemflow volume-weighted mean concentrations and fluxes for three size classes of broadleaf trees and three size classes of palms. The concentrations of most solutes were higher in stemflow than in rainfall and increased with increasing tree and palm size. Concentration enrichments from rainfall to stemflow and throughfall were particularly high (81-fold) for NO3-. Stemflow fluxes of NO3- and H+ exceeded throughfall fluxes but stemflow fluxes of other solutes were less than throughfall fluxes. Stemflow solute fluxes to the forest soil were dominated by fluxes on babassu palms, which represented only 4\% of total stem number and 10\% of total basal area. For NO3-, stemflow contributed 51\% of the total mass of nitrogen delivered to the forest floor (stemflow + throughfall) and represented more than a 2000-fold increase in NO3- flux compared what would have been delivered by rainfall alone on the equivalent area. Because these highly localized fluxes of both water and NO3- persist in time and space, they have the potential to affect patterns of soil moisture, microbial populations and other features of soil biogeochemistry conducive to the creation of hotspots for nitrogen leaching and denitrification, which could amount to an important fraction of total ecosystem fluxes. Because these hotspots occur over very small areas, they have likely gone undetected in previous studies and need to be considered as an important feature of the biogeochemistry of palm-rich tropical forest.}, language = {en} } @article{BaeseElsenbeerNeilletal.2012, author = {B{\"a}se, Frank and Elsenbeer, Helmut and Neill, Christopher and Krusche, Alex V.}, title = {Differences in throughfall and net precipitation between soybean and transitional tropical forest in the southern Amazon, Brazil}, series = {Agriculture, ecosystems \& environment : an international journal for scientific research on the relationship of agriculture and food production to the biosphere}, volume = {159}, journal = {Agriculture, ecosystems \& environment : an international journal for scientific research on the relationship of agriculture and food production to the biosphere}, publisher = {Elsevier}, address = {Amsterdam}, issn = {0167-8809}, doi = {10.1016/j.agee.2012.06.013}, pages = {19 -- 28}, year = {2012}, abstract = {The expansion of soybean cultivation into the Amazon in Brazil has potential hydrological effects at local to regional scales. To determine the impacts of soybean agriculture on hydrology, a comparison of net precipitation (throughfall, stemflow) in undisturbed tropical forest and soybean fields on the southern edge of the Amazon Basin in the state of Mato Grosso is needed. This study measured throughfall with troughs and stemflow with collar collectors during two rainy seasons. The results showed that in forest 91.6\% of rainfall was collected as throughfall and 0.3\% as stemflow, while in soybean fields with two-month old plants, 46.2\% of rainfall was collected as throughfall and 9.0\% as stemflow. Hence, interception of precipitation in soybean fields was far greater than in intact forests. Differences in throughfall, stemflow and net precipitation were found to be mainly associated with differences in plant structure and stem density in transitional forest and soybean cropland. Because rainfall interception in soybean fields is higher than previously believed and because both the area of cropland and the frequency of crop cycles (double cropping) are increasing rapidly, interception needs to be reconsidered in regional water balance models when consequences of land cover changes are analyzed in the Amazon soybean frontier region. Based on the continued expansion of soybean fields across the landscape and the finding that net precipitation is lower in soy agriculture, a reduction in water availability in the long term can be assumed.}, language = {en} }