Thèse – Systèmes d’élevage sur les prairies indigènes d’Amérique du Sud : quand la production rencontre la conservation

Pablo Modernel

Le mercredi 12 décembre 2018 à 13h30.

Generaal Foulkesweg 1. Gebouw 362. Postcode: 6703BG Wageningen, Pays-Bas

Vous pouvez suivre la soutenance via le lien :
https://weblectures.wur.nl/P2G/Player/Player.aspx?id=aulatv .

B. Dumont - INRA, France (rapporteur)
G. Bellocchi - INRA, France (rapporteur)
P.C. Struik - WUR, Pays-Bas (examinateur)
R. Ripoll-Bosch - WUR, Pays-Bas (examinateur)
P. Tittonell, INTA - Argentine (directeur de thèse, WUR)
M. Corbeels - CIRAD, Kenya (directeur de thèse, GAIA)
W. Rossing - WUR,Pays-Bas (co-encadrant)
S. Dogliotti - Universidad de la República, Uruguay (co-encadrant)

The livestock industry faces the double challenge of coping with the increasing demand for animal protein and reducing its high load on the environment. South American Río de la Plata grasslands comprise more than 500,000 km 2 , representing a hotspot of biodiversity. The livestock production systems of the region have relied on the coexistence of cattle and sheep on native grasslands for more than 400 years. However, high stocking rates have generated overgrazing, leading to low productivity levels and low income for farmers, and native grasslands are being replaced by non-native pastures and crops. The aim of this thesis was to explore options for increasing the sustainability of the native grassland-based livestock systems of the Rio de la Plata grasslands region. In Chapter 2 a literature review regarding the provision of ecosystem services by the Rio de la Plata grasslands and the land-use change processes of the last twenty years was conducted. The grasslands provide feed for 43 million heads of cattle and 14 million sheep. They store 5% of the total soil organic carbon stock of Latin America on 3% of the area, and they protect soils from erosion. The biome is habitat of 4000 native plant species, 300 species of birds, 29 species of mammals, 49 species of reptiles and 35 species of amphibians. Within the region, cropping areas and livestock herds have evolved in the last 20 years, with in certain areas the specialization in cropping and the use of feedlots and in other areas the occurrence of increased stocking rates, which could aggravate overgrazing problems. The most documented impacts of land-use change and overgrazing include: reduced soil organic carbon stocks and lower diversity of plants, birds and mammals, and increased soil erosion. We discussed how the transformation of grazing regimes to higher forage allowance regimes can contribute to greater meat production and enhanced ecosystem services from native grasslands. In Chapter 3 the economic and environmental performances of the current beef productionsystems of the Río de la Plata grasslands region were assessed. The chapter includes a beef farm typology of the whole region and the identification of positive deviant farms, which performed outstandingly in economic and environmental terms when compared to the other farms in the region, and, more generally, to farms from OECD countries. These farms achieved 192 kg LW ha -1 yr -1 or 201 US$ ha -1 year -1 with negligible fossil energy consumption and phosphorus surplus, low carbon footprint (13 kg CO 2 eq kg LW -1 ) and with over 95% of their land under native grassland. This means that the native grassland-based farming systems of the Río de la Plata grasslands region have the potential to produce high-quality beef with low use of inputs and at the same time can preserve biodiversity. In Chapter 4, the relationship between management and resilience to drought at paddock and farm level was studied. At paddock level, higher herbage allowance (kg herbage DM kg -1 animal LW) increased the resistance of herbage accumulation rate and animal weight to drought. A positive relationship was found between pre-drought herbage height and resistance of herbage accumulation to drought, which means that herbage height can be used as a guideline for grassland management. The sheep to cow ratio was negatively correlated with pregnancy rate of cows, a key variable defining meat production and farm income. In Chapter 5 a dynamic whole-farm simulation model (PASpALuM) is presented and evaluated, with the ultimate aim to aid farmers' decision making for better management of grassland systems . The model elucidates the relationships between grazing management, productivity and environmental impact. The herbage dynamics module was evaluated against experimental data, showing acceptable simulation of the seasonal dynamics of herbage height and mass. A simulation exercise explored the effect of grazing management on enteric methane emissions and soil organic carbon from 2007 to 2009.