This paper presents a new approach of spatiotemporally visualizing the simulation output of migratory insect dynamics and resultant vegetation changes in real-time. The visualization is capable of displaying simulated ecological phenomena in an intuitive manner, which allows research results to be easily understood by a wide range of users. In order to design a fast and efficient visualization technique, a simplified mathematical model is applied to intelligibly represent migrating groups of insects. In addition, impostors are used to accelerate rendering processes. The presented visualization method is implemented in an integrated spatiotemporal analysis system, which models, simulates and analyzes ecological phenomena such as insect migration through time at a variety of spatial resolutions.
The use of multi-layer perceptrons (MLP) to determine the relative significance of climatic variables to the establishment of insect pest species is described. Results show that the MLP are able to learn to accurately predict the establishment of a pest species within a specific geographic region. Analysis of the MLP yielded insights into the contribution of the individual input variables and allowed for the identification of those variables that were most significant in either encouraging or inhibiting establishment.
In Piedmont (Italy) the environmental changes due to human impact have had profound effects on rivers and their inhabitants. Thus, it is necessary to develop practical tools providing accurate ecological assessments of river and species conditions. We focus our attention on Salmo marmoratus, an endangered salmonid which is characteristic of the Po river system in Italy. In order to contribute to the management of the species, four different approaches were used to assess its presence: discriminant function analysis, logistic regression, decision tree models and artificial neural networks. Either all the 20 environmental variables measured in the field or the 7 coming from feature selection were used to classify sites as positive or negative for S. marmoratus. The performances of the different models were compared. Discriminant function analysis, logistic regression, and decision tree models (unpruned and pruned) had relatively high percentages of correctly classified instances. Although neither tree-pruning technique improved the reliability of the models significantly, they did reduce the tree complexity and hence increased the clarity of the models. The artificial neural network (ANN) approach, especially the model built with the 7 inputs coming from feature selection, showed better performance than all the others. The relative contribution of each independent variable to this model was determined by using the sensitivity analysis technique. Our findings proved that the ANNs were more effective than the other classification techniques. Moreover, ANNs achieved their high potentials when they were applied in models used to make decisions regarding river and conservation management.
Plant abundance data are often analysed using standard statistical procedures without considering their distributional features and the underlying ecological processes. However, plant abundance data, e.g. when measured in biodiversity monitoring programs, are often sampled using a hierarchical sampling procedure, and since plant abundance data in a hierarchical sampling procedure are typically both zero-inflated and over-dispersed, the use of a standard statistical procedure is sub-optimal and not the best possible practice in the modelling of plant abundance data. Two distributions (the zero-inflated generalised binomial distribution and the zero-inflated bounded beta distribution) are suggested as possible distributions for analysing either discrete, continuous, or ordinal hierarchically sampled plant cover data.
Statistical mechanics of relative species abundance (RSA) patterns in biological networks is presented. The theory is based on multispecies replicator dynamics equivalent to the Lotka–Volterra equation, with diverse interspecies interactions. Various RSA patterns observed in nature are derived from a single parameter related to productivity or maturity of a community. The abundance distribution is formed like a widely observed left-skewed lognormal distribution. It is also found that the “canonical hypothesis” is supported in some parameter region where the typical RSA patterns are observed. As the model has a general form, the result can be applied to similar patterns in other complex biological networks, e.g. gene expression.
The complexity of ecosystems is staggering, with hundreds or thousands of species interacting in a number of ways from competition and predation to facilitation and mutualism. Understanding the networks that form the systems is of growing importance, e.g. to understand how species will respond to climate change, or to predict potential knock-on effects of a biological control agent. In recent years, a variety of summary statistics for characterising the global and local properties of such networks have been derived, which provide a measure for gauging the accuracy of a mathematical model for network formation processes. However, the critical underlying assumption is that the true network is known. This is not a straightforward task to accomplish, and typically requires minute observations and detailed field work. More importantly, knowledge about species interactions is restricted to specific kinds of interactions. For instance, while the interactions between pollinators and their host plants are amenable to direct observation, other types of species interactions, like those mentioned above, are not, and might not even be clearly defined from the outset. To discover information about complex ecological systems efficiently, new tools for inferring the structure of networks from field data are needed. In the present study, we investigate the viability of various statistical and machine learning methods recently applied in molecular systems biology: graphical Gaussian models, L1-regularised regression with least absolute shrinkage and selection operator (LASSO), sparse Bayesian regression and Bayesian networks. We have assessed the performance of these methods on data simulated from food webs of known structure, where we combined a niche model with a stochastic population model in a 2-dimensional lattice. We assessed the network reconstruction accuracy in terms of the area under the receiver operating characteristic (ROC) curve, which was typically in the range between 0.75 and 0.9, corresponding to the recovery of about 60% of the true species interactions at a false prediction rate of 5%. We also applied the models to presence/absence data for 39 European warblers, and found that the inferred species interactions showed a weak yet significant correlation with phylogenetic similarity scores, which tended to weakly increase when including bio-climate covariates and allowing for spatial autocorrelation. Our findings demonstrate that relevant patterns in ecological networks can be identified from large-scale spatial data sets with machine learning methods, and that these methods have the potential to contribute novel important tools for gaining deeper insight into the structure and stability of ecosystems.
Environmental sensor networks are now commonly being deployed within environmental observatories and as components of smaller-scale ecological and environmental experiments. Effectively using data from these sensor networks presents technical challenges that are difficult for scientists to overcome, severely limiting the adoption of automated sensing technologies in environmental science. The Realtime Environment for Analytical Processing (REAP) is an NSF-funded project to address the technical challenges related to accessing and using heterogeneous sensor data from within the Kepler scientific workflow system. Using distinct use cases in terrestrial ecology and oceanography as motivating examples, we describe workflows and extensions to Kepler to stream and analyze data from observatory networks and archives. We focus on the use of two newly integrated data sources in Kepler: DataTurbine and OPeNDAP. Integrated access to both near real-time data streams and data archives from within Kepler facilitates both simple data exploration and sophisticated analysis and modeling with these data sources.
Standard interfaces for data and information access facilitate data management and usability by minimizing the effort required to acquire, catalog and integrate data from a variety of sources. The authors have prototyped several data management and analysis applications using Sensor Web Enablement Services, a suite of service protocols being developed by the Open Geospatial Consortium specifically for handling sensor data in near-real time. This paper provides a brief overview of some of the service protocols and describes how they are used in various sensor web projects involving near-real-time management of sensor data.
Ecological communities consist of a large number of species. Most species are rare or have low abundance, and only a few are abundant and/or frequent. In quantitative community analysis, abundant species are commonly used to interpret patterns of habitat disturbance or ecosystem degradation. Rare species cause many difficulties in quantitative analysis by introducing noises and bulking datasets, which is worsened by the fact that large datasets suffer from difficulties of data handling. In this study we propose a method to reduce the size of large datasets by selecting the most ecologically representative species using a self organizing map (SOM) and structuring index (SI). As an example, we used diatom community data sampled at 836 sites with 941 species throughout the French hydrosystem. Out of the 941 species, 353 were selected. The selected dataset was effectively classified according to the similarities of community assemblages in the SOM map. Compared to the SOM map generated with the original dataset, the community pattern gave a very similar representation of ecological conditions of the sampling sites, displaying clear gradients of environmental factors between different clusters. Our results showed that this computational technique can be applied to preprocessing data in multivariate analysis. It could be useful for ecosystem assessment and management, helping to reduce both the list of species for identification and the size of datasets to be processed for diagnosing the ecological status of water courses.
This paper presents a model of a population of error-prone self-replicative species (replicators) that interact with its environment. The population evolves by natural selection in an environment whose change is caused by the evolutionary process itself. For simplicity, the environment is described by a single scalar factor, i.e. its temperature. The formal formulation of the model extends two basic models of Ecology and Evolutionary Biology, namely, Daisyworld and Quasispecies models. It is also assumed that the environment can also change due to external perturbations that are summed up as an external noise. Unlike previous models, the population size self-regulates, so no ad hoc population constraints are involved. When species replication is error-free, i.e. without mutation, the system dynamics can be described by an (n + 1)-dimensional system of differential equations, one for each of the species initially present in the system, and another for the evolution of the environment temperature. Analytical results can be obtained straightforwardly in low-dimensional cases. In these examples, we show the stabilizing effect of thermal white noise on the system behavior. The error-prone self-replication, i.e. with mutation, is studied computationally. We assume that species can mutate two independent parameters: its optimal growth temperature and its influence on the environment temperature. For different mutation rates the system exhibits a large variety of behaviors. In particular, we show that a quasispecies distribution with an internal sub-distribution appears, facilitating species adaptation to new environments. Finally, this ecologically inspired evolutionary model is applied to study the origin and evolution of public opinion.
The study of functional structure in species assemblages emphasizes the detection of significant guild aggregation patterns. Thus, protocols based on intensive resampling of empirical data have been proposed to assess guild structure. Such protocols obtain the frequency distribution of a given functional similarity metric, and identify a threshold value (often the 95th percentile) beyond which clusters in a functional dendrogram are considered as significant guilds (using one-tailed tests). An alternative approach sequentially searches for significant differences between clusters at decreasing levels of similarity in a dendrogram until one is detected, then assumes that all subsequent nodes should also be significant. Nevertheless, these protocols do not test both the significance and sign of deviations from random at all levels of functional similarity within a dendrogram. Here, we propose a new bootstrapping approach that: (1) overcomes such pitfalls by performing two-tailed tests for each node in a dendrogram of functional similarity after separately determining their respective sample distributions, and (2) enables the quantification of the relative contribution of guild aggregation and functional divergence to the overall functional structure of the entire assemblage. We exemplify this approach by using long-term data on guild dynamics in a vertebrate predator assemblage of central Chile. Finally, we illustrate how the interpretation of functional structure is improved by applying this new approach to the data set available.
Precious ecological information extracted from limnological long-term time series advances the theory on functioning and evolution of freshwater ecosystems. This paper presents results of applications of artificial neural networks (ANN) and evolutionary algorithms (EA) for ordination, clustering, forecasting and rule discovery of complex limnological time-series data of two distinctively different lakes. Ten years of data of the shallow and hypertrophic Lake Kasumigaura (Japan) are utilized in comparison with 13 years of data of the deep and mesotrophic Lake Soyang (Korea). Results demonstrate the potential that: (1) recurrent supervised ANN and EA facilitate 1-week-ahead forecasting of outbreaks of harmful algae or water quality changes, (2) EA discover explanatory rule sets for timing and abundance of harmful outbreaks algal populations, and (3) non-supervised ANN provide clusters to unravel ecological relationships regarding seasons, water quality ranges and long-term environmental changes.
Ecological patterns are difficult to extract directly from vegetation data. The respective surveys provide a high number of interrelated species occurrence variables. Since often only a limited number of ecological gradients determine species distributions, the data might be represented by much fewer but effectively independent variables. This can be achieved by reducing the dimensionality of the data. Conventional methods are either limited to linear feature extraction (e.g., principal component analysis, and Classical Multidimensional Scaling, CMDS) or require a priori assumptions on the intrinsic data dimensionality (e.g., Nonmetric Multidimensional Scaling, NMDS, and self organized maps, SOM).
We compared the ability of three machine learning algorithms (linear discriminant analysis, decision tree, and support vector machines) to automate the classification of calls of nine frogs and three bird species. In addition, we tested two ways of characterizing each call to train/test the system. Calls were characterized with four standard call variables (minimum and maximum frequencies, call duration and maximum power) or eleven variables that included three standard call variables (minimum and maximum frequencies, call duration) and a coarse representation of call structure (frequency of maximum power in eight segments of the call). A total of 10,061 isolated calls were used to train/test the system. The average true positive rates for the three methods were: 94.95% for support vector machine (0.94% average false positive rate), 89.20% for decision tree (1.25% average false positive rate) and 71.45% for linear discriminant analysis (1.98% average false positive rate). There was no statistical difference in classification accuracy based on 4 or 11 call variables, but this efficient data reduction technique in conjunction with the high classification accuracy of the SVM is a promising combination for automated species identification by sound. By combining automated digital recording systems with our automated classification technique, we can greatly increase the temporal and spatial coverage of biodiversity data collection.
The Global Positioning System (GPS) has been increasingly used during the past decade to monitor the movements of free-ranging animals. This technology allows to automatically relocate fitted animals, which often results into a high-frequency sampling of their trajectory during the study period. However, depending on the objective of trajectory analysis, this study may quickly become difficult, due to the lack of well designed computer programs. For example, the trajectory may be built by several “parts” corresponding to different behaviours of the animal, and the aim of the analysis could be to identify the different parts, and thereby the different activities, based on the properties of the trajectory. This complex task needs to be performed into a flexible computing environment, to facilitate exploratory analysis of its properties. In this paper, we present a new class of object of the R software, the class “ltraj” included in the package adehabitat, allowing the analysis of animals' trajectories. We developed this class of data after an extensive review of the literature on the analysis of animal movements. This class of data facilitates the computation of descriptive parameters of the trajectory (such as the relative angles between successive moves, distance between successive relocations, etc.), graphical exploration of these parameters, as well a numerous tests and analyses developed in the literature (first passage time, trajectory partitioning, etc.). Finally, this package also contains numerous examples of animal trajectories, and a working example illustrating the use of the package.
We describe a semantic data validation tool that is capable of observing incoming real-time sensor data and performing reasoning against a set of rules specific to the scientific domain to which the data belongs. Our software solution can produce a variety of different outcomes when a data anomaly or unexpected event is detected, ranging from simple flagging of data points, to data augmentation, to validation of proposed hypotheses that could explain the phenomenon. Hosted on the Jena Semantic Web Framework, the tool is completely domain-agnostic and is made domain-aware by reference to an ontology and Knowledge Base (KB) that together describe the key resources of the system being observed. The KB comprises ontologies for the sensor packages and for the domain; historical data from the network; concepts designed to guide discovery of internet resources unavailable in the local KB but relevant to reasoning about the anomaly; and a set of rules that represent domain expert knowledge of constraints on data from different kinds of instruments as well as rules that relate types of ecosystem events to properties of the ecosystem. We describe an instance of such a system that includes a sensor ontology, some rules describing coastal storm events and their consequences, and how we relate local data to external resources. We describe in some detail how a specific actual event—an unusually high chlorophyll reading—can be deduced by machine reasoning to be consistent with being caused by benthic diatom resuspension, consistent with being caused by an algal bloom, or both.
Ecological systems are governed by complex interactions which are mainly
nonlinear. In order to capture this complexity and nonlinearity, statistical
models recently gained popularity. However, although these models are commonly
applied in ecology, there are no studies to date aiming to assess the
applicability and performance. We provide an overview for nature of the wide
range of the data sets and predictive variables, from both aquatic and
terrestrial ecosystems with different scales of time-dependent dynamics, and
the applicability and robustness of predictive modeling methods on such data
sets by comparing different statistical modeling approaches. The methods
considered k-NN, LDA, QDA, generalized linear models (GLM) feedforward
multilayer backpropagation networks and pseudo-supervised network ARTMAP. For
ecosystems involving time-dependent dynamics and periodicities whose frequency
are possibly less than the time scale of the data considered, GLM and
connectionist neural network models appear to be most suitable and robust,
provided that a predictive variable reflecting these time-dependent dynamics
included in the model either implicitly or explicitly. For spatial data, which
does not include any time-dependence comparable to the time scale covered by
the data, on the other hand, neighborhood based methods such as k-NN and ARTMAP
proved to be more robust than other methods considered in this study. In
addition, for predictive modeling purposes, first a suitable, computationally
inexpensive method should be applied to the problem at hand a good predictive
performance of which would render the computational cost and efforts associated
with complex variants unnecessary.
Lake Tuendae is a shallow, alkaline, artificially constructed Mojave Desert aquatic environment housing the endangered Mojave Tui Chub (Gila bicolar mohavensis). Detailed physiological response studies have been reported on the Mojave Tui Chub but few on the physico-chemical state of Lake Tuendae, one of four key Mojave Desert habitats for this species. Two sampling campaigns (spring 2004 and 2005) were conducted with correlation analysis, cluster analysis (CA) and principal component analysis (PCA) of physico-chemical water column and surface sediment quality parameters performed. CA proved useful in displaying parameter similarity for initial interpretation. PCA proved to be a more reliable display model and permitted the reduction of 14 parameters for the water column to four principal components accounting for 71% of the total variability. For surface sediments, four principal components accounted for 81% of the total variability. This work highlights the successful use of chemometric multivariate techniques in helping elucidate the physico-chemical make-up of shallow desert aquatic environments, and instructive for investigators assessing the health of aquatic species in such habitats.
The dynamics of the dissolved oxygen in water bodies is the result of complex interactions involving physical and biological processes. Understanding how the balance of these influences determines the amount of oxygen available for living organisms is a key factor to interpret the water body conditions, and eventually to use dissolved oxygen as an indicator of the water quality. In this paper we present a Qualitative Reasoning model developed to improve understanding of changes in the amount of dissolved oxygen in different segments of the river Mesta in Bulgaria. Effects on dissolved oxygen result from changes in physical, chemical and biological processes induced both by natural and anthropogenic activities within the watershed. To explore the possibility of establishing a landmark value that may change according to specific conditions, we developed the concept of flexible value mapping, which dynamically captures changes in the dependencies between the landmark value and the values of other quantities as the conditions of the system change during the simulations. The paper also discusses the concept of dominance of a specific process over other competing processes affecting a quantity. With the model described here, we aim to discuss possible solutions to interesting modelling problems and to provide the community of ecological modellers support for educational activities and water resources management.
The field of ecoinformatics is concerned with gaining a greater understanding of complex ecological systems. Many ecoinformatic tools, including artificial neural networks (ANNs), can shed important insights into the complexities of ecological data through pattern recognition and prediction; however, we argue that ecological knowledge has been used in a very limited fashion to shape the manner in which these approaches are applied. The present study provides a simple example of using ecological theory to better direct the use of neural networks to address a fundamental question in aquatic ecology—how are local stream macroinvertebrate communities structured by a hierarchy of environmental factors operating at multiple spatial scales? Using data for 195 sites in the western United States, we developed single-scale, multi-scale and hierarchical multi-scale neural networks relating EPT (Orders: Ephermeroptera, Plecoptera, Trichoptera) richness to environmental variables quantified at 3 spatial scales: entire watershed, valley bottom (100s–1000s m), and local stream reach (10s–100s m). Results showed that models based on multiple spatial scales greatly outperformed single-scale analyses (R = 0.74 vs. R¯ = 0.51) and that a hierarchical ANN, which accounts for the fact that valley- and watershed-scale drivers influence local characteristics of the stream reach, provided greater insight into how environmental factors interact across nested spatial scales than did the non-hierarchical multi-scale model. Our analysis suggests that watershed drivers play a greater role in structuring local macroinvertebrate assemblages via their direct effects on local-scale habitats, whereas they play a much smaller indirect role through their influence on valley-scale characteristics. For the hierarchical model, the strongest predictors of EPT richness included descriptors of climate, land-use and hydrology at the watershed scale, land-use at the valley scale, and substrate characteristics and riparian cover at the reach scale. In summary, our results highlight the importance of incorporating environmental hierarchies to better understand and predict local patterns of macroinvertebrate assemblage structure in stream ecosystems. More generally, our case study serves to emphasize how incorporating prior ecological knowledge into ANN model structure can strengthen the relevance of ecoinformatic techniques for the broader scientific community.
In a global assessment, canonical correspondence analysis (CCA) and partial CCA were used to ordinate Lake Huron phytoplankton abundances from June and August 1991 and environmental variables. June taxa were associated with NO3 and chloride, while August taxa were associated with SiO2 and temperature, and to some degree, with TSP and NH3. Dominant taxa were Asterionella formosa, Fragilaria capucina, Fragilaria crotonensis, Tabellaria fenestrata, and Urosolenia eriensis in June, and Achnanthidium minutissimum, Cyclotella #6, Cyclotella comensis, Cyclotella michiganiana, and Cyclotella pseudostelligera in August reflecting seasonal change. From local analysis using results from CCA and partial CCA in fuzzy relational analysis, A. minutissimum and C. comensis were influential in June, while F. crotonensis was influential in August. From linguistic translation and trophic status assignment, F. capucina and T. fenestrata indicated eutrophy, A. formosa indicated mesotrophy, C. pseudostelligera indicated mesotrophy–eutrophy, F. crotonensis and U. eriensis indicated oligotrophy–eutrophy, Cyclotella #6 indicated oligotrophy–mesotrophy, and C. michiganiana indicated oligotrophy. A linguistic solution with respect to trophic status is useful for policy makers and others interested in understanding water quality and ways to develop decisions about remediation.
Multivariate statistical analysis is a powerful method of examining complex datasets, such as species assemblages, that does not suffer from the oversimplification prevalent in many univariate analyses. However, identifying whether data points on a multivariate plot are clustered is subjective, as there is no determination of significant differences between the points and no indication of the level of confidence in those points. The validity of drawing such conclusions may therefore be considered suspect. This paper describes a method of bootstrapping calculated principal components to estimate a confidence radius, similar to confidence intervals in univariate techniques. Plotting 3D scatterplots of the principal components, with the size of the spherical point representative of the level of confidence of the estimate, gives a clear and visual indication of significant difference between the points — where the spheres overlap there is no significant difference. We apply the technique to mammal assemblages at sites in Epping Forest (Essex, UK) that differ in the level of disturbance present and find that differences between some sites that appear large using traditional principal components analysis are actually not significantly different at the 95% confidence level, while other sites do differ significantly. Sites that differ most in anthropogenic disturbance are not significantly different in terms of assemblage structure.
This paper presents a new statistical techniques — Bayesian Generalized Associative Functional Networks (GAFN), to model the dynamical plant growth process of greenhouse crops. GAFNs are able to incorporate the domain knowledge and data to model complex ecosystem. By use of the functional networks and Bayesian framework, the prior knowledge can be naturally embedded into the model, and the functional relationship between inputs and outputs can be learned during the training process. Our main interest is focused on the Generalized Associative Functional Networks (GAFNs), which are appropriate to model multiple variable processes. Three main advantages are obtained through the applications of Bayesian GAFN methods to modeling dynamic process of plant growth. Firstly, this approach provides a powerful tool for revealing some useful relationships between the greenhouse environmental factors and the plant growth parameters. Secondly, Bayesian GAFN can model Multiple-Input Multiple-Output (MIMO) systems from the given data, and presents a good generalization capability from the final single model for successfully fitting all 12 data sets over 5-year field experiments. Thirdly, the Bayesian GAFN method can also play as an optimization tool to estimate the interested parameter in the agro-ecosystem. In this work, two algorithms are proposed for the statistical inference of parameters in GAFNs. Both of them are based on the variational inference, also called variational Bayes (VB) techniques, which may provide probabilistic interpretations for the built models. VB-based learning methods are able to yield estimations of the full posterior probability of model parameters. Synthetic and real-world examples are implemented to confirm the validity of the proposed methods.
Coalbed Natural Gas extraction usually results in the production of excess, or product, water, necessitating a strategy for disposal and minimizing landscape and habitat impacts. In the Powder River Basin in Wyoming, product water is either discharged into ephemeral streams or retention/detention ponds. Monitoring these water bodies is important for environmental, habitat, and human health perspectives. This study assessed the benefits of using higher spatial resolution ASTER image, in contrast to more commonly used moderate-resolution Landsat imagery, for detecting smaller water bodies in the Powder River Basin. ASTER and Landsat Thematic Mapper (TM) images were acquired concomitantly and classified following similar methods to identify water bodies for three color classes and a range of sizes. Results showed that the ASTER image had significantly higher accuracies for detecting clear and green colored water bodies, but did not demonstrate significant improvement for detecting turbid water bodies. ASTER also showed significant improvements in detecting small-scale water bodies. However this improved performance was somewhat offset due to the misclassification of other landscape elements as water in the ASTER image. Overall when compared to Landsat TM image, ASTER image more accurately detected more water bodies, especially those with a relatively small surface area, with the two images producing similar results at large scales. The application of ASTER is therefore appropriate for monitoring and evaluation of water bodies in the Powder River Basin and elsewhere.
The flow regulation of rivers, mainly for flood control in wet season and water supply in dry season, dramatically altered the hydrological regime in the downstream, thus imposed significant impacts on the aquatic ecosystem. The evolution of riparian vegetation is an important indicator to quantify these impacts. This research focuses on the understanding of the vegetation dynamics and succession of riparian zones due to flow regulations by reservoir operation. The study developed an integrated model which couples a two-dimensional hydrodynamics module with a vegetation evolution module. Owing to the ability to well present spatial heterogeneity and local interactions, the vegetation module applied a cellular automata approach. To more precisely describe the complex morphology and topography, and to improve computation efficiency, an unstructured cellular automata (UCA) scheme which implemented a triangular mesh was used. The developed model was applied to a typical compound channel of the Lijiang River, which has been largely affected by the flow regulations of the Qingshitan reservoir for navigation purpose. The model was calibrated by the historical vegetation data, the field observations and the controlling experiment data. Through the scenarios simulation, the effects of flow regulation on riparian vegetation dynamics were analyzed. In addition, the potentials of UCA in riparian vegetation modelling were well explored.