Search

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jenvman.2007.09.010 Byline: M.J. Aitkenhead (a), I.H. Aalders (b) Abstract: Modelling land cover change from existing land cover maps is a vital requirement for anyone wishing to understand how the landscape may change in the future. In order to test any land cover change model, existing data must be used. However, often it is not known which data should be applied to the problem, or whether relationships exist within and between complex datasets. Here we have developed and tested a model that applied evolutionary processes to Bayesian networks. The model was developed and tested on a dataset containing land cover information and environmental data, in order to show that decisions about which datasets should be used could be made automatically. Bayesian networks are amenable to evolutionary methods as they can be easily described using a binary string to which crossover and mutation operations can be applied. The method, developed to allow comparison with standard Bayesian network development software, was proved capable of carrying out a rapid and effective search of the space of possible networks in order to find an optimal or near-optimal solution for the selection of datasets that have causal links with one another. Comparison of land cover mapping in the North-East of Scotland was made with a commercial Bayesian software package, with the evolutionary method being shown to provide greater flexibility in its ability to adapt to incorporate/utilise available evidence/knowledge and develop effective and accurate network structures, at the cost of requiring additional computer programming skills. The dataset used to develop the models included GIS-based data taken from the Land Cover for Scotland 1988 (LCS88), Land Capability for Forestry (LCF), Land Capability for Agriculture (LCA), the soil map of Scotland and additional climatic variables. Author Affiliation: (a) Department of Plant and Soil Science, University of Aberdeen, St. Machar Drive, Aberdeen AB24 3UU, Scotland, UK (b) The Macaulay Institute, Craigiebuckler, Aberdeen AB15 8QH, Scotland, UK Article History: Received 12 April 2006; Revised 16 July 2007; Accepted 14 September 2007

Abstract: A neural network approach was used to predict the presence and concentration of a range of endocrine disrupting compounds (EDCs), based on field observations. Soil sample concentrations of endocrine disrupting compounds (EDCs) and site environmental characteristics, drawn from the National Soil Inventory of Scotland (NSIS) database, were used. Neural network models were trained to predict soil EDC concentrations using field observations for 184 sites. The results showed that presence/absence and concentration of several of the EDCs, mostly no longer in production, could be predicted with some accuracy. We were able to predict concentrations of seven of 31 compounds with r2 values greater than 0.25 for log-normalised values and of eight with log-normalised predictions converted to a linear scale. Additional statistical analyses were carried out, including Root Mean Square Error (RMSE), Mean Error (ME), Willmott's index of agreement, Percent Bias (PBIAS) and ratio of root mean square to standard deviation (RSR). These analyses allowed us to demonstrate that the neural network models were making meaningful predictions of EDC concentration. We identified the main predictive input parameters in each case, based on a sensitivity analysis of the trained neural network model. We also demonstrated the capacity of the method for predicting the presence and level of EDC concentration in the field, identified further developments required to make this process as rapid and operator-friendly as possible and discussed the potential value of a system for field surveys of soil composition. [Copyright &y& Elsevier]

A novel process-based soil model (MOSES – Modelling Soil Ecosystem Services) is described and tested using field data. The model is designed to provide information about a soil profile at approximately 1-cm depth resolution, on a 1min timestep. Conceptualisation of the model has targeted a set of soil ecosystem service-related functions, including carbon sequestration, water buffering and biomass productivity, with the model framework designed to allow inclusion of additional processes and functions over time. Processes implemented within the model include thermal conduction, water movement (using pedotransfer functions to determine hydraulic conductivity, matric potential and related parameters), organic matter pool dynamics and gas/solute diffusion. The organic matter status of the profile is initialised using iterative runs of the RothC model to determine partition sizes of Decomposable Plant Material, Resistant Plant Material and other carbon pools. The outputs of the model are used to evaluate soil ecosystem service provision. MOSES has been designed to allow the implementation of further soil processes in the future, with the intention of expanding the variety of soil ecosystem functions and services that can be modelled. Validation of the model against detailed time-series field measurements of CO2 concentration and emission, temperature and water content of a freely draining podzolic soil in Ireland have shown it to be effective at simulating these specific parameters, with statistically significant association between measured and modelled values of temperature at all depths, for saturation at shallow depths and for CO2 at all depths apart from the surface layer. Overall, the model was seen to perform better at shallow depth, with lower levels of accuracy in deeper layers. The model has also been shown capable of simulating soil functional provision, based on specific parameters and processes. The need to add additional processes to MOSES in order to improve the simulation of soil ecosystem service provision, multifunctionality and the effects of external drivers such as climate change and management is discussed. [Copyright &y& Elsevier]

Abstract: Effective land cover mapping often requires the use of multiple data sources and data interpretation methods, particularly when no one data source or interpretation method provides sufficiently good results. Method-oriented approaches are often only effective for specific land cover class/data source combinations, and cannot be applied when different classification systems or data sources are required or available. Here we present a method, based on Endorsement Theory, of pooling evidence from multiple expert systems and spatial datasets to produce land cover maps. Individual ‘experts’ are trained to produce evidence for or against a class, with this evidence being categorised according to strength. An evidence integration rule set is applied to evidence lists to produce conclusions of different strength regarding individual classes, and the most likely class identified. The only expert system design implemented currently within the methodology is a neural network model, although the system has been designed to accept information from decision trees, fuzzy k-means and Bayesian statistics as well. We have used the technique to produce land cover maps of Scotland using three classification systems of varying complexity. Mapping accuracy varied between 52.6% for a map with 96 classes to 88.8% for a map with eight classes. The accuracy of the maps generated is higher than when individual datasets are used, showing that the evidence integration method applied is suitable for improving land cover mapping accuracy. We showed that imagery was not necessarily the most important data source for mapping where a large number of classes are used, and also showed that even data sources that produce low accuracy scores when used for mapping by themselves do improve the accuracy of maps produced using this integrative approach. Future work in developing the method is identified, including the inclusion of additional expert systems and improvement of the evidence integration, and evaluation carried out of the overall effectiveness of the approach. [Copyright &y& Elsevier]

Spatial arrangement can be an important factor affecting competition among plants. We evaluated three ways to improve the effectiveness of angular dispersion (AD) for describing spatial arrangement in plant neighbourhood models. First, we modified AD formula by weighting each neighbour by its competitive influence. We calculated this using two different competition indices to derive an AD of competitive influence, rather than of equally weighted plant locations, around a subject plant. Secondly, we constrained the effect of AD on the neighbourhood model using an optimised parameter that defines the minimum value that AD can adopt. Thirdly, we included the direction in which competition is concentrated (the mean azimuth of the weighted AD) in the growth models. These developments were evaluated within a radial growth model of Scots pine and birch growing in semi-natural, spatially heterogeneous forest. Weighted AD resulted in significant improvements in predicted radial growth of target trees over the traditional measure of AD. The optimised parameter that defines the minimum value of AD consistently evolved values significantly higher than zero. This suggests that clumped and dispersed neighbourhoods do not differ in their negative effects on a subject tree as much as expected. The inclusion of directional components of the weighted AD did not improve the accuracy of the growth models. Weighting of the angular dispersion of neighbours improved the performance of local competition models. [Copyright &y& Elsevier]

Abstract: Decision tree classification provides a rapid and effective method of categorising datasets. Many algorithmic methods exist for optimising decision tree structure, although these can be vulnerable to changes in the training dataset. An evolutionary method is presented which allows decision tree flexibility through the use of co-evolving competition between the decision tree and the training data set. This method is tested using two different datasets and gives results comparable with or superior to other classification methods. A final discussion argues for the utility of decision trees over algorithmic or other alternative methods such as neural networks, particularly in situations where a large number of variables are being considered. [Copyright &y& Elsevier]

A wide variety of watershed-scale attributes can be used as predictors of the export of dissolved organic carbon (DOC) from a watershed. However, the complexity and number of relationships makes the development of generally applicable mechanistic models for prediction of DOC export based on measurement of factors difficult. Here we have applied neural network modelling methods to the prediction of stream flux and daily DOC export from several watersheds of varying size within the Dee valley, in north-east Scotland. A two-stage process was carried out in which first a model was developed which used a large number of variables thought to be relevant to DOC export, and then the possibility of using a restricted set of variables was investigated in order to reduce the amount of analysis required in order to produce accurate DOC export predictions. The results showed that it is possible to predict DOC export using input variables corresponding broadly to the factors responsible for soil formation, and that a single sample site may provide enough information to allow prediction for an entire watershed. However, in order to achieve a model with statistically significant results, it is necessary to use multiple sample sites per watershed, and to use measured rather than modelled flow values. Discussion is made of the effectiveness of the neural network method in developing models of DOC export, and of problems with the method (particularly in the inability to use NN models for process-based models). [Copyright &y& Elsevier]

Abstract: Modelling land use change is often constrained by imperfect and incomplete data sources. This paper explores three modelling methodologies and their ability to predict agricultural land use on the basis of information from the Scottish Agricultural Census. This dataset, which contains information on ownership, land use and employment statistics for the majority of Scotland, is restricted by law concerning the level of detail which can be provided, and as such is both the best available source of information for agricultural practice in Scotland and is partial and incomplete. It is demonstrated that the methodologies applied to the problem (neural network, Bayesian network and decision tree), with a limited number of relevant drivers included in the modelling process, are capable of use for the prediction of changes in land use, suitable for policy analysis. The reasons for selecting these particular modelling approaches included a need to deal with a large amount of noisy, inaccurate data, and the fact that each is capable of successfully investigating and quantifying unknown relationships between dataset variables. The greatest success, measured as a combination of accuracy, data-handling flexibility and ease of model comprehension by the user, was achieved by the decision tree method. [Copyright &y& Elsevier]

We describe an approach in which a neural network (NN) can be trained on sets of driving variables (inputs) and output variables relating to spatial structure. The trained NN provides a predictive tool for defining spatial structure in a specified ecological system. We demonstrate the approach using a modified version of the Crawley and May [J. Theor. Biol. 125 (1987) 475] model that describes simplified annual/perennial plant interactions in a disturbed system. The model is implemented within a cellular automaton with randomised start conditions and is run for periods of up to 50 time steps (50 years). Neural networks are trained using a large set of modelled situations, and the ability of the network to predict plant spatial distributions is then measured. Different image analysis methods are applied to the plant array, and the ability of each method to provide an accurate description of the end-state of the modelled system is investigated. Reconstruction of the plant array from these image analysis measurements is carried out using a stochastic error minimisation method. The partial derivatives method is applied to the trained neural network in order to determine which variables input to the model most strongly influence the eventual plant population distribution. In the example presented, it is found that a relatively simple boundary:area ratio measurement provides a rapid and effective method of describing the spatial structure of the plant community, while the variables that are most influential on the system’s end-state are those describing annual fecundity and perennial mortality rates. [Copyright &y& Elsevier]

A convincing argument has been made in the literature for the advantages gained by using modular neural networks (NN) instead of homogeneous structures. Here, a modular NN design was used in conjunction with evolutionary algorithm methods to evolve an animat capable of learning behavioural patterns at several levels of complexity. A parallel was drawn between the training of the animat and the stages of learning experienced by the young of many animals. Following movement learning, the animat was eventually capable of navigating an environment and avoiding obstacles. Discussion is made of how such an animat with many degrees of freedom can develop complex behavioural patterns. [Copyright &y& Elsevier]

A neural network based facial recognition program (FADER—FAce DEtection and Recognition) was developed and tested. The hardware and software components were all standard commercial design, allowing the system to be built for minimal cost. Using a set of 1000 face and 1000 ‘no-face’ images, we achieved 94.7% detection rate, and a 0.6% false positive rate. Three different neural network models were applied to face recognition, using single images of each subject to train the system. A novel adaptation of the Hebbian connection strength adjustment model gave the best results, with 74.1% accuracy achieved. Each of the system''s components, including an intermediate substructure detection network, was subject to evolutionary computation in order to optimise the system performance. [Copyright &y& Elsevier]

Soil, streamwater and climatic variables were measured hourly over several month periods in two situations in North-East (NE) Scotland, using data loggers and other measuring instruments. One of the locations was on agricultural land near Inverness while the other was at an acidic peat moorland site in the River Dee catchment. The data sets were used to train neural networks using three different methods, including a novel, biologically plausible system. Temporal pattern recognition capabilities using each method were investigated. The novel method proved equally capable in predicting future variable values using large data sets as the other two methods. An argument is made for this method, termed the ‘Local Interaction’ method, providing valid competition to other neural network and statistical methods in the detection of patterns and prediction of events in complex biological systems. [Copyright &y& Elsevier]

Recently, much research has been done on the possibilities of autonomous robots navigating and performing obstacle avoidance. Vision is an important method of gathering information about obstacles and other objects within an environment. A virtual environment was created, in which an animat (virtual organism) navigates using a visual system comparable to that used by biological organisms. A spike-based neural network model was applied to learning. Various applicable rules on network topology were used, and examination was made of how well the animat learnt under varying sizes of visual input layer and intermediate ‘processing’ layer. In addition, three different training regimes were applied and their different merits discussed. It was discovered that increasing layer size in general improved the performance of the animat, provided that the sizes of the input array and the processing arrays correspond. The learning curve of the animat over time was investigated, allowing an optimal training time to be determined. These findings allow an insight into how well such a system can perform and how its design and training may be optimized. [Copyright &y& Elsevier]

Abstract: In this study we explored the effects of numerical optimisation on the ability of three competition indices (Lorimer, Heygi and Pukkala–Kolström) to predict the annual radial growth of Scots pine (Pinus sylvestris L.) in Glen Affric, Scotland. We present methods that combined neural networks and simulated annealing to determine the optimal specification for each index. Simulating annealing was used to discover the optimal specification of a series of competition indices, while a neural network was used to discover the best-fit model with a given competition index. Other inputs to the neural network model included tree DBH (diameter at breast height), environmental (exposure, soil wetness) and topographic (aspect, altitude) variables. The neural network prediction accuracy was fed back to the optimisation algorithm, which altered the specification of the competition index based on changes in the prediction accuracy. The neural network design itself was also subjected to simulated annealing to obtain an optimised predictive model. Validation of the neural network model was carried out using data not used to train the model. The optimisation approach decreased the error rate for the neural network models. It was also shown (a) that the Pukkala–Kolström competition index model gave a markedly better prediction of tree growth rates, and (b) that the angle of inclusion of neighbouring trees in calculating competition indices should be lower than is often suggested. [Copyright &y& Elsevier]

• Development and proof-of-concept for a new method of soil pore network extraction. • Soil networks analysed using metrics from network science and transport geography. • Networks also used as environment for agent-based model of simulated consumers. • As expected, Cambisol networks were larger, more structured than for Arenosols. • Soil type, interaction with parameters both affected model populations outcomes. Quantifying soil structural and ecological heterogeneity is crucial for understanding their interactions and their relationships to the resilience and health of the wider ecosystem. However, a clear understanding of how structural heterogeneity affects soil biodiversity is still emerging. Previous work has primarily used expensive, often laboratory-based methods to quantify soil pore network structure, and typically separated study of structural and biological dimensions. Here, we test whether standard network metrics can be used to quantify structural heterogeneity in soil pore networks, and how this network structure, along with characteristics of the consumer and resource populations, affects the heterogeneity of a population of consumers. Specifically, we extract simplified soil pore networks from digital photographs of soil profiles and apply established metrics from network science and transport geography to quantify and compare the networks. The networks are also used as the medium for an agent-based model of generalised consumers, to analyse the effects of consumer and resource parameterisations and network structure. Combining network analysis and simulation modelling in this way can provide insights on the structure, function, and diversity possible in the soil, as well as avenues for exploring the impact of future structural or environmental changes. [ABSTRACT FROM AUTHOR]