Jean-François Guégan

Artificial neural networks as a tool in ecological modelling, an introduction

Abstract Artificial neural networks (ANNs) are non-linear mapping structures based on the function of the human brain. They have been shown to be universal and highly flexible function approximators for any data. These make powerful tools for models, especially when the underlying data relationships are unknown. In this reason, the international workshop on the applications of ANNs to ecological modelling was organized in Toulouse, France (December 1998). During this meeting, we discussed different methods, and their reliability to deal with ecological data. The special issue of this ecological modelling journal begins with the state-of-the-art with emphasis on the development of structural dynamic models presented by S.E. Jorgensen (DK). Then, to illustrate the ecological applications of ANNs, examples are drawn from several fields, e.g. terrestrial and aquatic ecosystems, remote sensing and evolutionary ecology. In this paper, we present some of the most important papers of the first workshop about ANNs in ecological modelling. We briefly introduce here two algorithms frequently used; (i) one supervised network, the backpropagation algorithm; and (ii) one unsupervised network, the Kohonen self-organizing mapping algorithm. The future development of ANNs is discussed in the present work. Several examples of modelling of ANNs in various areas of ecology are presented in this special issue.

Ecology drives the worldwide distribution of human diseases.

Identifying the factors underlying the origin and maintenance of the latitudinal diversity gradient is a central problem in ecology, but no consensus has emerged on which processes might generate this broad pattern. Interestingly, the vast majority of studies exploring the gradient have focused on free-living organisms, ignoring parasitic and infectious disease (PID) species. Here, we address the influence of environmental factors on the biological diversity of human pathogens and their global spatial organization. Using generalized linear multivariate models and Monte Carlo simulations, we conducted a series of comparative analyses to test the hypothesis that human PIDs exhibit the same global patterns of distribution as other taxonomic groups. We found a significant negative relationship between latitude and PID species richness, and a nested spatial organization, i.e., the accumulation of PID species with latitude, over large spatial scales. Additionally, our results show that climatic factors are of primary importance in explaining the link between latitude and the spatial pattern of human pathogens. Based on our findings, we propose that the global latitudinal species diversity gradient might be generated in large part by biotic interactions, providing strong support for the idea that current estimates of species diversity are substantially underestimated. When parasites and pathogens are included, estimates of total species diversity may increase by more than an order of magnitude.

Energy availability and habitat heterogeneity predict global riverine fish diversity

Processes governing patterns of richness of riverine fish species at the global level can be modelled using artificial neural network (ANN) procedures. These ANNs are the most recent development in computer-aided identification and are very different from conventional techniques,. Here we use the potential of ANNs to deal with some of the persistent fuzzy and nonlinear problems that confound classical statistical methods for species diversity prediction. We show that riverine fish diversity patterns on a global scale can be successfully predicted by geographical patterns in local river conditions. Nonlinear relationships, fitted by ANN methods, adequately describe the data, with up to 93 per cent of the total variation in species richness being explained by our results. These findings highlight the dominant effect of energy availability and habitat heterogeneity on patterns of global fish diversity. Our results reinforce the species-energy theory and contrast with those from a recent study on North American mammal species, but, more interestingly, they demonstrate the applicability of ANN methods in ecology.

International Variability of Ages at Menarche and Menopause: Patterns and Main Determinants

AbstractThe purpose of this study was to review published studies on the variability of age at menarche and age at menopause throughout the world, and to identify the main causes for age variation in the timing of these events. We first present a summary table including mean (or median) values of the age at menarche in 67 countries, and of the age at menopause in 26 countries. General linear models showed that mean age at menarche was strongly linked to the mean female life expectancy, suggesting that one or several variables responsible for inequalities in longevity similarly influenced the onset of menarche. A closer examination of the data revealed that among several variables reflecting living conditions, the factors best explaining the variation in age at menarche were adult illiteracy rate and vegetable calorie consumption. Because adult illiteracy rate has some correlation with the age at which children are involved in physical activities that can be detrimental in terms of energy expenditure, our results suggest that age at menarche reflects more a trend in energy balance than merely nutritional status. In addition, we found the main determinant of age at menopause to be the mean fertility. This study thus suggests that, on a large scale, age at menarche is mainly determined by extrinsic factors such as living conditions, while age at menopause seems to be mainly influenced by intrinsic factors such as the reproductive history of individuals. Finally, these findings suggest that human patterns cannot be addressed solely by traditional, small-scale investigations on single populations. Rather, complementary research on a larger scale, such as this study, may be more appropriate in defining some interesting applications to the practical problems of human ecology.

Time-dependent spectral analysis of epidemiological time-series with wavelets.

In the current context of global infectious disease risks, a better understanding of the dynamics of major epidemics is urgently needed. Time-series analysis has appeared as an interesting approach to explore the dynamics of numerous diseases. Classical time-series methods can only be used for stationary time-series (in which the statistical properties do not vary with time). However, epidemiological time-series are typically noisy, complex and strongly non-stationary. Given this specific nature, wavelet analysis appears particularly attractive because it is well suited to the analysis of non-stationary signals. Here, we review the basic properties of the wavelet approach as an appropriate and elegant method for time-series analysis in epidemiological studies. The wavelet decomposition offers several advantages that are discussed in this paper based on epidemiological examples. In particular, the wavelet approach permits analysis of transient relationships between two signals and is especially suitable for gradual change in force by exogenous variables.

Climate drives the meningitis epidemics onset in west Africa.

Background Every year West African countries within the Sahelo-Sudanian band are afflicted with major meningococcal meningitis (MCM) disease outbreaks, which affect up to 200,000 people, mainly young children, in one of the worlds poorest regions. The timing of the epidemic year, which starts in February and ends in late May, and the spatial distribution of disease cases throughout the “Meningitis Belt” strongly indicate a close linkage between the life cycle of the causative agent of MCM and climate variability. However, mechanisms responsible for the observed patterns are still not clearly identified. Methods and Findings By comparing the information on cases and deaths of MCM from World Health Organization weekly reports with atmospheric datasets, we quantified the relationship between the seasonal occurrence of MCM in Mali, a West African country, and large-scale atmospheric circulation. Regional atmospheric indexes based on surface wind speed show a clear link between population dynamics of the disease and climate: the onset of epidemics and the winter maximum defined by the atmospheric index share the same mean week (sixth week of the year; standard deviation, 2 wk) and are highly correlated. Conclusions This study is the first that provides a clear, quantitative demonstration of the connections that exist between MCM epidemics and regional climate variability in Africa. Moreover, this statistically robust explanation of the MCM dynamics enables the development of an Early Warning Index for meningitis epidemic onset in West Africa. The development of such an index will undoubtedly help nationwide and international public health institutions and policy makers to better control MCM disease within the so-called westward–eastward pan-African Meningitis Belt.

The use of artificial neural networks to assess fish abundance and spatial occupancy in the littoral zone of a mesotrophic lake

The present work describes a comparison of the ability of multiple linear regression (MLR) and artificial neural networks (ANN) to predict fish spatial occupancy and abundance in a mesotrophic reservoir. Models were run and tested with 306 observations obtained by the sampling point abundance method using electrofishing. For each of the 306 samples, the relationships between physical parameters and the abundance and spatial occupancy of various fish species were studied. For the 15 fish species occurring in the lake, six main fish populations were retained to perform comparisons between ANN and MLR models. Each of the six MLR and ANN models had eight independent environmental variables (i.e. depth, distance from the bank, slope of the bottom, flooded vegetation cover, percentage of boulders, percentage of pebbles, percentage of gravel and percentage of mud) and one dependent variable (fish density for the considered population). To determine the population assemblage, principal component analysis (PCA) was performed on the partial coefficients of the MLR and on the relative contribution of each independent variable of ANN models (determined using Garsons algorithm). The results stress that ANN are more suitable for predicting fish abundance at the population scale than MLR. In the same way, a higher level of ecological complexity, i.e. community scale, was reliably obtained by ANN whereas MLR presented serious shortcomings. These results show that ANN are an appropriate tool for predicting population assemblage in ecology.

A nested parasite species subset pattern in tropical fish: host as major determinant of parasite infracommunity structure

The number of monogenean gill parasite species associated with fish hosts of different sizes is evaluated for 35 host individuals of the West African cyprinid Labeo coubie. The length of host individuals explains 86% of the total variation in monogenean species richness among individuals. Larger hosts harbour more species than smaller ones. The existence of a hierarchical association of parasite species in individuals of L. coubie is demonstrated. Monogenean infracommunities on larger fish hosts consist of all species found on smaller hosts plus those restricted to the larger size categories, suggesting some degree of compositional persistence among host individuals. The findings provide strong support for an interpretation of the relationship between monogenean parasite species richness and host body size in terms of a nested species subset pattern, thus providing a new record of repetitive structure and predictability for parasite infracommunities of hosts.

Can host body size explain the parasite species richness in tropical freshwater fishes

SummaryThe variability of monogenean gill ectoparasite species richness in 19 West African cyprinid species was analyzed using the following seven predictor variables: host size, number of drainage basins, number of sympatric cyprinid species, host diversity, association with mainland forest, host ecology, and monogenean biological labelling. The size of the host species accounted for 77% of the variation in the number of parasite species per host, and host ecology an additional 8%. Together the effects of host size and host ecology accounted for 85% of the variation in monogenean species richness. This study shows that the deciding factors for explaining monogenean species richness in West African cyprinid fishes are host species size and host ecology. These results were compared with main factors responsible for parasite species richness in fish communities. Other possible explanations of monogenean community structure in west African cyprinids are discussed.

Water-borne transmission drives avian influenza dynamics in wild birds: the case of the 2005-2006 epidemics in the Camargue area.

Transmission and persistence of avian influenza viruses (AIV) among wildlife remains an unresolved issue because it depends both on the ecology of the host (e.g. population density, migration) and on the environment (e.g. AIV persistence in water). We have developed a mathematical model that accounts for both AIV epidemics and bird community dynamics. The model is parameterized using bird counts and AIV prevalence data. Results suggest that the transmission patterns driving the dynamics of infection at our study site (Camargue, South of France) involved both a density-dependent and a water-borne transmission processes. Water-borne transmission is, however, the main determinant of the disease dynamics and observed prevalence level. This pattern of transmission highlights the importance of the persistence of viral particles in water in AIV dynamics in wild birds.