Rock fragments modify soil properties, and can be a potential reservoir of water. Besides, recent studies showed that this coarse soil fraction is chemically active, release nutrients, and could therefore be involved in biogeochemical nutrient cycles. However, these studies do not answer the question whether the coarse soil fraction has significant nutritive functions. This present work attempted to assess if rock fragments may act as water and nutrient sources for poplar saplings, a deciduous species. A green house experiment was conducted using remoulded soils built by mixing fine earth and pebbles (2 cm < diameter of the rock fragments < 5 cm) at three percentages: 0 (control), 20, and 40% in volume, and either with calcareous or inert pebbles (quartz). Both fine earth and calcareous pebbles were collected from the Ap horizon of a calcareous lacustrine limestone silty soil. One cutting of Populus robusta was planted in each pot (10 replicates per modality), saturated once then irrigated to maintain a moderate water stress. Growth and evapotranspiration were followed, while water stress status was measured by stomatal conductivity every day during two drying periods of 10 days. After three months, biomass and cations (Ca, Mg, K) were measured. Results showed that pebbles can participate to plant nutrition, and reduced water stress, though the dilution effect was the main effect on plant development: the presence of pebbles reduced the growth. Nevertheless, plants suffered less from water stress when grown with calcareous pebbles, especially at 40% pebbles, compared to modalities with quartz pebbles. In addition, mineralomass of plants grown with 20% calcareous pebbles were similar to the control, and higher than that of modalities with quartz pebbles. In addition, plants had a biomass 26% higher when grown with calcareous than with quartz pebbles. These results indicate that plants access nutrients from pebbles and that growth conditions were significantly better in pots with calcareous compared to quartz pebbles at moderate proportion (20%). For modalities with 40% pebbles, no difference between the types of pebbles was found most probably because of the detrimental effects of rock fragments limiting root development. This study strengthen the hypothesis that coarse soil fraction may act as water and nutrient sources. The concept of an inert stone matrix that, from the plant point of view, only dilutes ecological functions of the soil, must be revised.