Possibility Theory: A New Approach to Uncertainty Analysis? (3 pp)

Background, Aims and Scope The problem of the evaluation of practitioner's belief and belief-related uncertainties on LCA results obtained from different methodological choices has been addressed so far by scenario modeling, Cultural Theory perspectives and probabilistic simulation. The direct evaluation of belief and related uncertainties could be of interest, e.g. when the information available (resulting from classical uncertainty analysis or the application of the precautionary principle) do not allow one to choose between methodological alternatives leading to different LCA results and conclusions. The difficulty of modeling belief arises from the additive nature of classical measures, e.g. probabilities. Since the 1960s, non-additive measures (e.g. possibilities) have been developed and applied to model belief in real world problems. The aim of this paper is to discuss the application of possibility measures in LCA for uncertainty analysis in complement to classical approaches. Methods The nature and the meaning of possibilities are briefly introduced by comparison with probabilities (subjective or not) in order to enlighten strengths, drawbacks and complementarities. A tentative possibilistic approach based on the evaluation of a posteriori possibilities of final LCA results depending on a priori possibilities of the methodological choices behind the calculations is described, also by means of an application example. Results and Outlook. A new approach for the modeling of practitioner's belief and belief-related uncertainties in complement of classical methods of uncertainty analysis has been proposed for discussion. Uncertainty can be characterized by confidence intervals and indexes that could help practitioners in making methodological choices and could improve the interpretation and reliability of LCA results, still increasing its sophistication.     

A biogeographic delineation of the European Alpine System based on a cluster analysis of Carex curvula-dominated grasslands

Biogeographic delineations within the European temperate mountains remain poorly understood, as there has been little effort to assemble and analyze vegetation relevés covering Pyrenees, Alps, Carpathians and Balkans altogether. Our study tackles this issue by focusing on the widely distributed alpine acidic grasslands dominated by Carex curvula. Cluster analysis of more than 800 vegetation relevés revealed the European-scale spatial patterns of vascular plant diversity in these alpine grasslands. The geographical distribution of floristic clusters was partly congruent with the physiography of European mountains. Southern European ranges (Southern Balkans and Pyrenees) exhibit a high level of endemism and corresponding floristic clusters are well separated from the others. Marked floristic similarities between the Easternmost Alps, the Carpathians, and the Northern Balkans (Stara Planina) supported a major floristic boundary that runs through the Austrian Alps and that is likely the legacy of a shared Quaternary history. Within the Alps, floristic clustering was mainly driven by ecological drivers and not geography. This paper presents the first detailed study of spatial patterns of species distribution within the European Alpine System, based on a comprehensive analysis of within- and between-community species diversity. It shows that the quantitative analysis of large and consistent data sets may question the traditional delineations of biogeographic regions within European mountains.     

Estimating plant responses to climate by direct gradient analysis and geographic distribution analysis

We characterised the climatic behaviour of 53 woody species in terms of the climatic factors that play the main role in controlling species distribution in the study area. Floristic and climatic data were obtained from 150 stands in sites under climatic control (i.e. eu-climatopes). The sampling strategy used allowed a reliable match between floristic and climatic observations. Different methods of frequency analysis and goodness-of-fit tests were used to identify associations between species occurrence and climatic characteristics. The species' responses were summarised by statistics describing ecological preferences and amplitudes, and species were grouped accordingly. A Gaussian response model was fitted to the abundance data along the main climatic gradients for selected species and response surfaces were derived by spatial analysis for a set of indicator species. Frequency analysis methods detected 42 indicator taxa for the Baudiere's Qe drought index, and lower numbers, 34 and 22, respectively, for the mean minimum coldest-month temperature and the daily temperature range in the coldest month. Goodness-of-fit tests revealed a lower number of ecological profiles with statistically significant deviations from equidistribution. We discuss the relative performance of the different methods and suggest that the combined use of statistical tests and frequency analyses may improve estimation of the environmental requirements of species. We also recommend using the species' responses to key environmental factors as reliable criteria in the definition of plant functional types. This revised version was published online in August 2006 with corrections to the Cover Date.     

Occurrence patterns of butterflies (Rhopalocera) in semi-natural pastures in southeastern Sweden

Occurrence patterns of butterflies (Rhopalocera) were analysed in relation to locality characteristics, and nested patterns were evaluated, based on a survey of 100 semi-natural pastures in the county of Östergötland, southeastern Sweden. Species richness of butterflies was positively related with locality-size, but not with the density of pastures in the surroundings. Species richness was lower in heavily grazed pastures compared with localities with a low or intermediate grazing pressure. This suggests that a uniformed, heavy grazing pressure should be avoided if butterflies are to be protected, even though such a management is favourable for many vascular plants. Out of 45 analysed species, 73% exhibited a significantly nested pattern, and species regarded to be sedentary had generally a more nested pattern. Several butterflies (such as Erynnis tages, Mellicta athalia and Pyrgus malvae) are possible to use as indicators of a relatively intact butterfly community. In a short-term perspective, the great bulk of butterflies confined to semi-natural pastures would be saved if management were adapted to the requirements of butterflies only at the localities richest in species. However, for conservation strategies to be successful over a longer time, whole landscapes harbouring the majority of the species and with a high density of semi-natural pastures should be selected and be given priority for conservation.     

Determination of diagnostic species with statistical fidelity measures

Abstract. Statistical measures of fidelity, i.e. the concentration of species occurrences in vegetation units, are reviewed and compared. The focus is on measures suitable for categorical data which are based on observed species frequencies within a vegetation unit compared with the frequencies expected under random distribution. Particular attention is paid to Bruelheide's u value. It is shown that its original form, based on binomial distribution, is an asymmetric measure of fidelity of a species to a vegetation unit which tends to assign comparatively high fidelity values to rare species. Here, a hypergeometric form of u is introduced which is a symmetric measure of the joint fidelity of species to a vegetation unit and vice versa. It is also shown that another form of the binomial u value may be defined which measures the asymmetric fidelity of a vegetation unit to a species. These u values are compared with phi coefficient, chi‐square, G statistic and Fisher's exact test. Contrary to the other measures, phi coefficient is independent of the number of relevés in the data set, and like the hypergeometric form of u and the chi‐square it is little affected by the relative size of the vegetation unit. It is therefore particularly useful when comparing species fidelity values among differently sized data sets and vegetation units. However, unlike the other measures it does not measure any statistical significance and may produce unreliable results for small vegetation units and small data sets. The above measures, all based on the comparison of observed/expected frequencies, are compared with the categorical form of the Dufrêne‐Legendre Indicator Value Index, an index strongly underweighting the fidelity of rare species. These fidelity measures are applied to a data set of 15 989 relevés of Czech herbaceous vegetation. In a small subset of this data set which simulates a phytosociological table, we demonstrate that traditional table analysis fails to determine diagnostic species of general validity in different habitats and large areas. On the other hand, we show that fidelity calculations used in conjunction with large data sets can replace expert knowledge in the determination of generally valid diagnostic species. Averaging positive fidelity values for all species within a vegetation unit is a useful approach to measure quality of delimination of the vegetation unit. We propose a new way of ordering species in synoptic species‐by‐relevé tables, using fidelity calculations.     

Indicators of seabird reproductive performance demonstrate the impact of commercial fisheries on seabird populations in the North Sea

In a world of growing anthropogenic pressures on biodiversity, effective indicators need to be specific and sensitive to the pressures in the ecosystem concerned, yet be simple enough to be interpreted by non-experts and straightforward enough to facilitate routine monitoring. Globally, seabirds are under increasing pressure as a result of anthropogenic activities and environmental variation. Traditionally, seabird indicators have been based on abundance at breeding colonies. However, as many species do not reach sexual maturity for several years, and may not attend the colony over this time period, such indicators may fail to capture the ecological complexity of the system concerned.We constructed two indicators of the state of nine seabird species that breed along the UK coast of the North Sea: (i) abundance of seabirds at breeding colonies, and (ii) probability of seabird breeding failure. The indicators were significantly and strongly correlated with each other for eight out of nine species, but the abundance indicator typically lagged the indicator on seabird breeding failure by two to three years. We then considered a third indicator which compared kittiwake (Rissa tridactyla) breeding success to the levels expected given the underlying environmental conditions; changes in the abundance indicator also lagged this by three years. We investigate how sensitive each of these indicators was to the impacts of fishing. We found that the species which had seen the greatest increases in breeding failure rate over the study period were those species which were most sensitive to fisheries pressure.By focussing on demographic parameters, and correcting for the underlying environmental conditions, we can detect potentially important population level changes at an earlier stage than by focussing on abundance alone. These indicators are able to more accurately capture the complexity of the ecosystem concerned and can be readily interpreted by policy-makers.     

The impact of site conditions and seed dispersal on restoration success in alluvial meadows

Abstract. We studied the restoration success of flood plain meadows in the northern Upper Rhine valley, where between 1988 and 1992, 35 ha of arable land was converted into grassland and subsequently managed for nature conservation. Remnant populations of typical alluvial meadow species were found in old meadows and along drainage ditches that dissect the whole area. We analysed the site conditions and phytosociological relevés in old and new meadows. Small differences in site parameters between old and new meadows contrasted with a clear floristic differentiation between the two meadow types. The vegetation of old meadows was much more differentiated along prevailing environmental gradients than the vegetation of new meadows. Despite the favourable site conditions for the re‐establishment of species‐rich meadows on the former arable land, restoration success was limited to the vicinity of remnant stands. In contrast to old meadows, indicator species of new grassland were still typical species of regularly disturbed ruderal and arable habitats, often capable of building up a persistent seed bank. The precise mapping of 23 target species revealed that even wind dispersal predominantly leads to re‐establishment in the close circumference of parent plants. We found no indication that regular flooding, hay‐making and autumnal grazing had an impact on recolonization of newly created grassland. Even under favourable conditions for the re‐establishment of target species, restoration success in alluvial meadows proved to be strongly dispersal limited. We discuss the implications of our findings for future restoration management in grasslands.     

An improved approach to evaluate car sharing options

We develop an improved approach to evaluate car sharing options under uncertain environments with the combination of Fuzzy Analytic Hierarchy Process (F-AHP) and Fuzzy Technique for Order Preference by Similarity to Ideal Solution (F-TOPSIS), which consists of three steps. In the first step, we propose a SCUMN (Specific, Comprehensive, Understandable, Measurable, and Neutral) methodology to identify appropriate indicators and obtain a final list of 24 indicators according to their relevance to car sharing options. In the second step, we determine the weight of each indicator with F-AHP and conduct consistency check of the comparison matrix of selected indicators. In the third step, comparison of different options is performed with selected indicators and F-TOPSIS. A case study is provided to validate the proposed approach. Twenty-four indicators are identified to evaluate five different car sharing options and rank them according to their closeness coefficients in decreasing order. And thirty-one sensitivity analysis experiments are conducted to figure out the influence of indicators on decision making. The experimental results show that the proposed approach is capable of evaluating car sharing options with uncertainty and vagueness. F-AHP is able to determine the weight for each selected indicator and F-TOPSIS demonstrates its advantage in comparing potential options.