No description available.

La contribution des changements d’usage des sols aux émissions de gaz à effet de serre d’origine anthropique est estimée à 17% pour la décennie 2000, en grande partie liée à la déforestation. L’un des facteurs principaux de ces changements est l’expansion des terres agricoles pour les besoins locaux de développement, mais également sous l’effet des exportations stimulées par la mondialisation. Pour cette raison, des préoccupations nouvelles surgissent quant aux effets des politiques sur l’usage des sols par le biais des marchés internationaux. Ce travail présente trois illustrations concrètes où ces effets peuvent être d’ampleur conséquente : i) l’intensification de l’agriculture dans les pays en voie de développement, ii) les accords commerciaux, et iii) les politiques d’agrocarburants. Les résultats montrent que pour chacune de ces politiques, les réponses des marchés sont susceptibles de jouer un rôle déterminant dans le bilan des gaz à effet de serre. L’atténuation du changement climatique par l’intensification des cultures conduit à des réductions d’émissions, mais l’effet rebond de la demande pourrait annuler une part substantielle des bénéfices attendus sur les surfaces de terres cultivées. L’exemple d’un possible accord entre l’Union européenne et le Mercosur montre les effets négatifs que peut induire la libéralisation de certains produits agricoles, si des mesures d’accompagnement adéquates ne sont pas mises en place. Enfin, l’effet des changements indirects d’affectation des sols est susceptible d’effacer une part substantielle des réductions d’émissions alléguées aux agrocarburants. Les réponses de l’affectation des sols aux différentes politiques dépendent néanmoins de nombreux paramètres comportementaux, et il est difficile d’en fournir une estimation chiffrée précise. Plusieurs approches de modélisation sont utilisées ici pour quantifier ces effets et explorer les intervalles de confiance découlant des estimations actuelles de la littérature économétrique. La prise en compte de cette externalité dans l’évaluation des politiques publiques nécessite des approches nouvelles intégrant mieux les différents niveaux d’incertitude sur ces effets. / Land use change is estimated to have generated 17% of anthropogenic greenhouse gas emissions in the 2000s, a large part coming from deforestation. The main driver of these emissions is expansion of agricultural activities, for the need of local development in tropical regions. However, they have also been caused by the dynamics of globalisation which has stimulated agricultural trade flows. Thus, today, there are new concerns with respect to how agricultural policies are influencing land use changes in other parts of the world through international market responses. In this work I consider three concrete illustrations of where these effects can be of significant magnitude: i) agriculture intensification in developing countries, ii) trade agreements, and iii) biofuel policies. I find that for each of these policies, market responses are likely to play a significant role in the final greenhouse gas emission balance. Mitigation of emissions through agricultural intensification could have quite beneficial outcomes, but the rebound effect on the demand side would offset a large part of greenhouse gas emission savings attributable to the land sparing effect. With the example of a possible EU-MERCOSUR trade agreement, I also show the adverse effect of liberalising certain specific agricultural products closely connected to land use change dynamics without adequate accompanying measures. Last, the indirect land use change effect of biofuels is likely to offset a large part of their alleged GHG emission savings. Land use change responses depend on many behavioural parameters, however, and providing precise estimates constitutes a challenge. I use different modelling approaches to quantify their magnitude and extensively explore the level of confidence on the basis of current state of econometric findings.New approaches should be elaborated to take account of this externality in public policy assessments, together with an appropriate consideration of the uncertainty ranges associated with these effects.

Questa tesi di dottorato analizza l’utilizzo del sorgo (Sorghum bicolour (L.) Moench) al fine di produrre energia elettrica, tramite combustione diretta della biomassa. Il focus della tesi è stato quello di sottolineare i benefici ed i rischi associati all’uso di tre genotipi di sorgo caratterizzati da diversa lunghezza del ciclo culturale (precoce, medio-tardivo e tardivo). La dinamica e la durata del processo di essicazione in campo sono state simulate attraverso un modello ad hoc (“sorghum haying model”), il quale integrato a CropSyst, è stato utilizzato per realizzare un’analisi del rischio produttivo stimando le perdite di biomassa (respirazione e meccanizzazione), ed i mancati affienamenti. Nell’analisi del rischio vengono stimati il numero di ettari necessari e la probabilità di eccedere la soglia di 64.000 ton ss anno-1 necessari per l’alimentazione di una centrale nell’Oltrepò pavese . Inoltre uno studio di Life Cycle Assessment è stato condotto per la valutazione dell’impatto ambientale dell’utilizzo del sorgo integrato a quello della paglia per il completamento del fabbisogno totale della centrale 94.000 ton ss anno-1. Particolare attenzione inoltre è stata data alla variazione del contenuto di C organico del suolo dovuto alla rimozione della paglia ed all’interramento dei mancati affienamenti di sorgo. Il genotipo precoce mostra le migliori performance produttive ed energetiche oltre che la più alta probabilità di eccedere la soglia di 64.000 ton ss anno-1. Lo studio di LCA non ha mostrato differenze significative tra i genotipi anche se il minor impatto ambientale, è stato evidenziato dal genotipo tardivo conseguenza dell’interramento della più alta quantità di mancati affienamenti. / This PhD thesis explores the use of sorghum (Sorghum bicolour (L.) Moench) as a dedicated bio-energy crop and highlights the benefits and risks associated with the use of early, medium-late and late sorghum genotypes to generate electricity by direct combustion in a biomass power plant. The dynamics and duration of the field drying process were simulated through the development of a specific model ("sorghum haying model"), which integrated with CropSyst, was used to perform a production risk assessment analysis estimating the biomass losses (respiration and mechanical), the haymaking failures and consequently to quantify the amount of dry baled biomass available for the power plant. In addition, the number of hectares needed to plant sorghum and the probability to exceed the threshold of 64000 Mg DM y-1, necessary to feed a biomass power plant in Oltrepò Pavese, were estimated. A complete Life Cycle Assessment (LCA) study was carried out in order to evaluate the environmental impact of the three sorghum genotypes involved in this study. The LCA study takes into consideration the use of winter wheat straw as an additional biomass source to satisfy the total biomass power plant needs (94000 Mg DM y-1). Particular attention was given to the soil organic C change (ΔSOC) due to straw removal and haymaking failures soil incorporation. Early genotype showed the best biomass production and energy performance as well as the highest probability to exceed the threshold of 64000 Mg DM y-1. The LCA results did not show significant differences between genotypes although the lower environmental impact, has been achieved by the late genotype due to the highest amount of haymaking failures incorporated in the soil.