Michel Sebillotte

The role of grain number, nitrogen nutrition and stem number in limiting pea crop (Pisum sativum) yields under agricultural conditions

In a small area of the Paris basin in France, 71 pea (Pisum sativum) crops, cv. Solara, were examined to elucidate the determinants of their yield over 3 years. The yields ranged from between 3.71 and 7.81 t ha−1, and averaged 5.85, 6.25 and 6.47 t ha−1 in 1988, 1989 and 1990, respectively. The large differences observed each year among fields were essentially related to grain number, whereas mean grain weight was less variable. Grain number and pod number were highly correlated. A substantial effect of nitrogen nutrition, assessed by a nitrogen nutrition index on grain number, was observed. The role of reproductive stem number on grain number was also examined, showing a threshold at 115 stems m−2, below which, under the conditions examined, large grain numbers per stem did not compensate for low stem numbers. Furthermore, we also show that branching might not be sufficient to compensate for low plant numbers in some crops when nitrogen nutrition is inadequate, and this also leads to a decrease in grain number.

Les fondements épistémologiques de l'évaluation des recherches tournées vers l'action

Abstract The epistemological foundations of researches turned toward the actions assessment. In the present context, characterised by an increasing demand for research from private and public decision makers, public research must adapt itself. In the cases of researches ‘for’ and ‘about’ regional development, this is traducing by a double requirement: to form pluridisciplinary collectives of researchers, and also to set up a specific partnership shape, melting numerous researchers gathered in the same geographical area. But, within the research institutions, these researches because it get off the usual canons, are putting the question of its assessment. To my opinion, scientific character is based on the way research are made more than the disciplinary character of the object. This article aims to show that assessment can not be narrowed to the only results and that it is necessary to evaluate how this researches are carried out. Doing researches ‘for’ the regional development implies researcher in the innovation processes which lead to add a third epistemological model, the one of the research-action, to the two models of the laboratory and of the terrain. Those researches are lead by multidisciplinary collectives of researchers in partnership with external of research. The system created allows to build in common from the partners problems the questions that have to be dealt and then, to build objects likely to be dealt scientifically by multidisciplinary collectives. Epistemologically speaking, it presupposes to construct rules that allows to build models. These models will have to organise, in principle, the contributions of different disciplines and to product knowledge scientifically allowing the partners to tackle problems, that lead them to appeal to research, more efficiently.