Enrico Feoli

Evaluation of environmental degradation in northern Ethiopia using GIS to integrate vegetation, geomorphological, erosion and socio-economic factors

The scale of human induced land degradation is very high in northern Ethiopia. However, there are only a few studies of land degradation in Ethiopia which provide an integrated assessment of the driving forces and consequences. A pilot study was made in Adwa district (Tigray) with the objective of evaluating the factors related to environmental degradation and assessing the effects of agricultural activities on the natural resources. Environmental data, a digital terrain model, vegetation, geomorphological, and erosion maps of the study area were integrated with socio-economic variables using a geographical information system (GIS). The classification of the data used in the integration, and the information extracted using the GIS gave three main groups of Tabias and four main groups of variables. The relationship among the Tabias (smallest administrative units), groups of Tabias and environmental variables were quantified using various statistical, multivariate numerical methods and fuzzy set analysis. The application of fuzzy set theory showed that each group of Tabia, recognized as a rural subsystem (RS) was associated with a group of variables. Analysis of variance showed that the three RSs were significantly different in most of the environmental variables considered. Human and livestock population densities, geomorphology, altitude and some natural and anthropogenic vegetation types were highly discriminatory. Percentage cover of evergreen scrub, bushlands and severe badlands are considered key indicators of the scale of environmental degradation. Results of the canonical correlation analysis (CCA) indicated that human pressure had more impact on the physiognomy of the vegetation than on its floristic composition. The Evergreen scrub vegetation type appeared to be expanding with increasing human influence signifying a decrease in biomass of vegetation as a result of collecting wood for fuel and other domestic uses. Bushland appeared to be expanding with the same trend. Woodland, which was abundant in one of the RSs, is considered to be the natural physiognomic vegetation type in the area and efforts to maintain it and/or recover it elsewhere in the study area are recommended. Food production in the area did not match population growth. The expected crop yield under good soil and rainfall conditions was low by any standard. The per capita energy obtained from the grain harvested in the farm plots ranged from 428 to 4347 cal per person per day (the average is far below the basic minimum required for mere subsistence) suggesting that there is a need for substantial supplementary income from off-farm activities and/or other sources. This study indicated that regional and site specific approaches and interaction with the people at all stages of the research, program development and implementation are required for sustainable rehabilitation.

Application of a spatial decision support system (SDSS) to reduce soil erosion in northern Ethiopia

A spatial decision support system (SDSS) based on multi-criteria and multi-objective decision analysis is applied in a case study in Ethiopia to reduce soil erosion on the basis of reallocation of crops according to their capacity to protect the soil. The case study is carried out in the Adwa district. The SDSS has been implemented using the widespread GIS software IDRISI 32 (release 2) and with the direct involvement of local stakeholders in defining factors and constraints. These are based on land cover-land use, altitude, potential erosion, proximity to roads, water and the relative soil protective capacity of each crop species. A reduction of soil loss from an average of 4.5 t ha−1 yr−1 to values below the risk threshold of soil degradation (1 t ha−1 yr−1) would be achieved through the application of the SDSS results. The biggest impediment to the reallocation exercise, however, is the shortage of cultivable land suitable for cultivation.

Syntaxonomical analysis of beech woods in the Apennines (Italy) using the program package IAHOPA

IAHOPA, an overlay program package performing intersection analysis and information analysis has been applied to a large data set of relevés of beech woods in the Apennines (Italy) completed by several authors following the Braun Blanquet approach. The results have been treated by several numerical methods testing classification efficiency and predictivity. Ecological indicator values have been used to test for predictivity. The classification proposed by Gentile has been confirmed in its main lines. However 2 new associations (Polysticho-Fagetum and Digitali-Fagetum) and 12 new subassociations are described. Furthermore the Veronico-Fagetum Montacchini 1972 has been recognized also for the Apennines. The clusters corresponding to the association level could be classified in two main alliances: Geranio nodosi-Fagion and Geranio striati-Fagion as suggested by Gentile, however their syntaxonomical justification should be based on numerical comparisons of the data from the entire area of European beech woods.

Partitioning an elevation gradient of vegetation from southeastern Ethiopia by probabilistic methods

Vegetation and environment have been analyzed along an altitudinal gradient in Harena Forest, Bale Mountains National Park, southeastern Ethiopia. Vegetation data include numbers of each tree and shrub species and cover-abundance values of each herbaceous species. Environmental data comprise edaphic factors, altitude and topography. The two vegetation layers data were analysed separately. Probabilistic similarity coefficients were computed between the relevés, and these values were used in subsequent computations for classification and ordination. Two sets of stratocoena, comprising 6 types each, derived on the basis of separate analyses of tree-shrub and herb layers of the forest were recognised. A combination of the two sets of stratocoena produced a total of 11 vegetation types. Environmental fuzzy set analysis was applied to determine the strength of the relationship of the relevés to the environmental factors. Autocorrelation analysis was applied to the eigenvectors of probabilistic similarity matrices and environmental data. Altitude appears to be more important thant the other environmental factors in controlling the zonation of the forest. Other important environmental influences on the vegetation include pH, organic matter content and texture of the soil. It is suggested that the whole forest be included in the National Park to create suitable conditions for adequate protection.

Modelling Environmental Responses of Plant Associations: A Review of Some Critical Concepts in Vegetation Study

The definition of vegetation types at different hierarchical levels, both to study the vegetation processes and for practical cartographic representation, is still considered a critical issue in many circles of plant ecologists. The problems are mainly related to the misleading idea that classification of the vegetation system, as developed by European phytosociologists during the last century within the discipline called syntaxonomy, would imply the assumption of the organismic concept of the plant community. After a short discussion on the role of Braun-Blanquet approach in plant ecology and in landscape ecology, the methods to detect multispecies responses along environmental gradients are briefly reviewed. In the main part of this article, we intend to stress that concepts considered critical, such as plant association and its ecological niche, are just operational tools that have nothing to do with the individualistic or organismic interpretation of plant communities in vegetation studies. Important to our views on vegetation, we believe that plant associations as well as the higher syntaxa can be regarded as fuzzy sets in an operational context for describing vegetation along ecological gradients in synthetic ways and can further the understanding of vegetation variation.

A numerical phytosociological study of the summits of the majella massive (Italy)

A phytosociological study of Majellas summits by the use of multivariate methods has allowed to discover the following new associations:Gnaphalio-Plantaginetum atratae on humocarbonatic soils in snowy depressions,Leontopodio-Elynetum on rendzina in wind exposed and stony places,Saxifrago-Papaveretum julici andCrepidi-Leontodontetum montani on protorendzina with gravel.Gnaphalio-Plantaginetum atratae andLeontopodio-Elynetum might be classified inElyno-Seslerietea, although the former presents character species ofSalicetea herbaceae; Saxifrago-Papaveretum julici andCrepidi-Leontodontetum montani are classified inThlaspidion stylosi (foeder. nova) ofThlaspidietea. Comparisons between some clustering methods and between ordination methods based on the mathematics of principal component analysis are discussed. The efficiency of sum of squares agglomeration when the plant communities vary in a continuum and the importance of a linearizing transformation in principal component analysis are stressed. Durch pflanzensoziologische Studien der Majella-Gipfel mittels Multivarianzmethoden wurden die vier folgenden Assoziationen neu festgestellt:Gnaphalio-Plantaginetum atratae auf Humuskarbonatböden in Schneetälchen, Leontopodio-Elynetum auf Rendzina in windexponierten Stellen,Saxifrago-Papaveretum julici undCrepidi-Leondontetum montani auf Protorendzina mit Schutt.Gnaphalio-Plantaginetum atratae undLeontopodio-Elynetum könnten zu denElyno-Seslerietea gestellt werden, obwohl die zuerst genannte Gesellschaft Charakterarten derSalicetea herbaceae enthält;Saxifrago-Papaveretum julici undCrepidi-Leontodontetum montani werden zum neuen VerbandThlaspidion stylosi derThlaspidietea gestellt. Vergleiche zwischen einigen Schwarmmethoden (clustering methods) und Ordinationsmethoden, die auf den mathematischen Grundlagen der ‘prineipal component’ Analyse beruhen, werden diskutiert. Die Effizienz der ‘sum of squares agglomeration’ bei Variation der Pflanzengemeinschaften in einem Kontinuum und die Wichtigkeit der linearisierenden Transformation bei der ‘principal component’ Analyse werden betont.

Processes of environmental degradation and opportunities for rehabilitation in Adwa, Northern Ethiopia

There are only a few studies of land cover-land use changes which provide an integrated assessment of the biophysical and societal causes and consequences of environmental degradation in Ethiopia. Our objectives were to determine the status of the environmental degradation, analyse and evaluate the relationships among vegetation, geomorphological and socio-economic factors contributing to environmental degradation, and propose opportunities for rehabilitation of these natural resources. Field and other environmental data in northern Ethiopia and those acquired by remote sensing techniques were used to accomplish these objectives. These were integrated with socio-economic data obtained from official sources using a Geographic Information System (GIS).Spatial information such as the percent of land cover-land use types and geomorphological categories, and the percent of each land cover-land use type in the geomorphological categories were calculated using Geographic Information System (GIS). The three most dominant features of the geomorphological categories (93.0%) are scarps and denuded rock slopes, erosion surfaces and badlands, while the three most dominant features in the land cover-land use types (71.3%) are croplands, open woodlands and bushlands. Badlands account for 38.7% of the geomorphological units and 41.8% of the croplands currently occur on badlands. Simple and partial correlation analyses were applied to explore the extent of the interaction between the anthropogenic and the natural system. The anthropogenic system is influenced by elevation, which is positively correlated with human population and livestock densities and area of croplands. The natural system finds its place only on steep slopes as shown by the positive correlation between woodland,slope, high potential erosion, scarps and denudational rock slopes. The study indicates that agriculture in the study area is in a critical environmental situation. A change of paradigm in land-use and development is needed to encourage participation of the landowners and users in the efforts to conserve the vegetation and the soil. This study provides sound options that could be used to rehabilitate the vegetation directly and to alleviate the current pressure on the land and improve human welfare indirectly. Matching the human and livestock densities with the carrying capacity of the land through recruitment of the surplus labour force for a modern economy, resettlement,off-farm employment and intensification of agriculture are t he long and short-term actions that may contribute to the rehabilitation of the degraded areas.

Ordination Based on Classification: Yet Another Solution?!

A new simple method of ordination is proposed. It is based on classification of releves, or variables, and matrix multiplication. An example of application is given with data from a simulated coenocline with different noise levels. The performance of the method may be considered as good as that of the most popular methods, or even better if the monotonous pattern is considered as an advantage.

Monitoring desertification risk with an index integrating climatic and remotely‐sensed data

In this paper a desertification risk index (DRI) based on the integration of climatic data and the normalised difference vegetation index (NDVI), obtained from National Oceanic Atmospheric Administration advance very high‐resolution radiometer (NOAA‐AVHRR) images, is discussed at the light of the aridity index and some eco‐physiological parameters. The good correlation between DRI, the aridity index and the eco‐physiological parameters suggests that DRI could be useful to measure the desertification risk. One advantage of DRI is that, with the help of a geographic information system (GIS), DRI maps can be easily obtained in short time and at relatively low costs.

On the resolving power of principal component analysis in plant community ordination

A comparison between centered and non-centered principal component analysis is made on the basis of the resolving power of the methods. The results indicate the appropriateness of using a centered PCA when the aim is an ordination of plant communities, and the noncentered PCA when the aim is to elicit the taxonomic structure of a collection.