Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → How to assess rangeland condition in semiarid ecosystems ? The indicative value of vegetation in the High Atlas Mountains, Morocco

Universität zu Köln (2009)

How to assess rangeland condition in semiarid ecosystems ? The indicative value of vegetation in the High Atlas Mountains, Morocco

Baumann, Gisela

Titre : How to assess rangeland condition in semiarid ecosystems ? The indicative value of vegetation in the High Atlas Mountains, Morocco

Auteur : Baumann, Gisela

Université de soutenance : Universität zu Köln

Grade : Doktorgrades der Mathematisch-Naturwissenschaftlichen 2009

The natural resource base of the world�s drylands is under continuous threat, particularly due to immense population growth, climate and land use change. Since the majority of these ecosystems are used as rangelands for domestic livestock, forage is (beneath drinking water) one of the most required ecosystem goods. For local land users as well as policy-makers it is crucial to quantify the actual supply of forage and try to predict potential limitations in the future. Ecological indicators as easily interpretable surrogates for complex ecological processes play an important role to assess rangeland condition. However, in highly stochastic arid and semiarid ecosystems it is still a challenge to identify reliable indicators detecting anthropogenic change against a background of natural variability. The present dissertation investigates three currently discussed rangeland indicators and their significance along a steep gradient of natural variability in southern Morocco. The research was conducted on four altitudinal levels along the southern slopes of the High Atlas Mountains. The transect represents a steep aridity gradient stretching from arid climate in the Basin of Ouarzazate to subhumid climate in the high mountain areas. First, the production (ANPP) and rain-use efficiency (RUE) of the vegetation was analyzed by means of an ecological field experiment comparing grazed, 1 year and 7 years rested vegetation. It revealed that ANPP and RUE are suitable and relatively �fast� indicators to quantify the actual supply of forage, but not to detect long-term and irreversible degradation processes. For that, the new parameters ANPPrel and RUErel are suggested, which standardize the production on the amount of initial biomass. In contrast to ANPP and RUE, they focus on the vitality of perennial forage plants which is altered on a longer time scale and allows a better comparison across ecosystems. Second, plant functional types (PFT) were tested as indicators. Particularly response groups and response traits were assembled to quantify grazing impact. In contrast to the initial goals of PFT research, those indicators were shown to be locally limited, since their predictive value was strongly influenced by resource stochasticity (aridity). It is deduced that using response groups and response traits for range assessment in arid and semiarid ecosystem is only reliable if their application is restricted to the local scale. Third, this work examines the local ecological knowledge of nomadic land-users in the research area. An interdisciplinary study among herdsmen of a local Berber fraction discovered that the �reliability� of plants and pastures functions as a local indicator which influences local range management. The �reliability�-concept integrates several spatial and temporal scales. It provides a new opportunity to quantify anthropological information and to compare it to ecological data. The present work depicts an important milestone for the application of ecological indicators in range assessment. It is essential for land users and policy-makers to choose the appropriate indicator level not only to economize costs and manpower, but also to reduce bias in indication processes. A combination of several hierarchically operating indicators is suggested for arid and semiarid ecosystems, for example measures of ANPP/RUE to quantify actual forage supply of pastures and ANPPrel/RUErel to detect areas affected by long-term degradation. Results of this work further build a data base for two different vegetation models. In the context of rapid climate and land use change we cannot afford further debates whether environmental changes have been caused by human impact or are just an impression of natural variability. My work shows the potential and limits of three ecological indicators under semiarid climate. It thus provides the framework for an appropriate application that allows an indication of human impact against a background of resource variability.

Mots clés : ecology , range assessment , indicator , Morocco , dryland, semi-arid


Version intégrale

Page publiée le 31 août 2010, mise à jour le 7 décembre 2017