Refine
Has Fulltext
- no (2)
Year of publication
- 2018 (2) (remove)
Document Type
- Article (2) (remove)
Language
- English (2) (remove)
Is part of the Bibliography
- yes (2)
Keywords
- Baltic Sea (1)
- Lolium perenne (1)
- Scirpus maritimus (1)
- hybrid model (1)
- model coupling (1)
- transient dynamics (1)
Institute
Time series of groundwater and stream water quality often exhibit substantial temporal and spatial variability, whereas typical existing monitoring data sets, e.g. from environmental agencies, are usually characterized by relatively low sampling frequency and irregular sampling in space and/or time. This complicates the differentiation between anthropogenic influence and natural variability as well as the detection of changes in water quality which indicate changes in single drivers. We suggest the new term "dominant changes" for changes in multivariate water quality data which concern (1) multiple variables, (2) multiple sites and (3) long-term patterns and present an exploratory framework for the detection of such dominant changes in data sets with irregular sampling in space and time. Firstly, a non-linear dimension-reduction technique was used to summarize the dominant spatiotemporal dynamics in the multivariate water quality data set in a few components. Those were used to derive hypotheses on the dominant drivers influencing water quality. Secondly, different sampling sites were compared with respect to median component values. Thirdly, time series of the components at single sites were analysed for long-term patterns. We tested the approach with a joint stream water and groundwater data set quality consisting of 1572 samples, each comprising sixteen variables, sampled with a spatially and temporally irregular sampling scheme at 29 sites in northeast Germany from 1998 to 2009. The first four components were interpreted as (1) an agriculturally induced enhancement of the natural background level of solute concentration, (2) a redox sequence from reducing conditions in deep groundwater to post-oxic conditions in shallow groundwater and oxic conditions in stream water, (3) a mixing ratio of deep and shallow groundwater to the streamflow and (4) sporadic events of slurry application in the agricultural practice. Dominant changes were observed for the first two components. The changing intensity of the first component was interpreted as response to the temporal variability of the thickness of the unsaturated zone. A steady increase in the second component at most stream water sites pointed towards progressing depletion of the denitrification capacity of the deep aquifer.
Resilience is a major research focus covering a wide range of topics from biodiversity conservation to ecosystem (service) management. Model simulations can assess the resilience of, for example, plant species, measured as the return time to conditions prior to a disturbance. This requires process-based models (PBM) that implement relevant processes such as regeneration and reproduction and thus successfully reproduce transient dynamics after disturbances. Such models are often complex and thus limited to either short-term or small-scale applications, whereas many research questions require species predictions across larger spatial and temporal scales. We suggest a framework to couple a PBM and a statistical species distribution model (SDM), which transfers the results of a resilience analysis by the PBM to SDM predictions. The resulting hybrid model combines the advantages of both approaches: the convenient applicability of SDMs and the relevant process detail of PBMs in abrupt environmental change situations. First, we simulate dynamic responses of species communities to a disturbance event with a PBM. We aggregate the response behavior in two resilience metrics: return time and amplitude of the response peak. These metrics are then used to complement long-term SDM projections with dynamic short-term responses to disturbance. To illustrate our framework, we investigate the effect of abrupt short-term groundwater level and salinity changes on coastal vegetation at the German Baltic Sea. We found two example species to be largely resilient, and, consequently, modifications of SDM predictions consisted mostly of smoothing out peaks in the occurrence probability that were not confirmed by the PBM. Discrepancies between SDM- and PBM-predicted species responses were caused by community dynamics simulated in the PBM and absent from the SDM. Although demonstrated with boosted regression trees (SDM) and an existing individual-based model, IBC-grass (PBM), our flexible framework can easily be applied to other PBM and SDM types, as well as other definitions of short-term disturbances or long-term trends of environmental change. Thus, our framework allows accounting for biological feedbacks in the response to short- and long-term environmental changes as a major advancement in predictive vegetation modeling.