In preliminary analyses, we have assessed some sources of variation in floral and nectar traits within a metapopulation of the insect-pollinated Pyrus bourgaeana in the Doñana National Park (SW, Spain). By considering five subpopulations (Reserva, Rocina, Matasgordas, Hato Ratón, and Hinojos) during two consecutive flowering seasons, it seems to emerge marked interpopulation and annual differences in both sorts of plant traits. For example, mean petal length varied significantly among populations (p < 0.05), with the longest and shortest petals occurring in Matasgordas and Vera, respectively. However, there was significant interaction between year and locality (p<0.0001), indicating that differences between localities were not consistent between years. Most interestingly, we have found temporally consistent tradeoffs between floral display and pollinator reward. For instance, whereas populations with large petals (e.g., Matasgordas) tended to have low amount of nectar, populations with small petals (e.g., Hato Ratón, Hinojos) had high amounts of nectar. These results suggest that P. bourgaeana invests resources either in attracting pollinator or in rewarding them. Nevertheless, our field observations, preliminary dendrochronological data, and published spatial analysis (Fedriani et al. 2010) suggest that P. bourgaeana trees clearly differ in age (range, 15-120 years), size and level of aggregation ? from a small scale (a few meters to hundred of meters) to a landscape scale (tens of kilometers). All those factors can potentially influence flower and nectar traits (Watkinson 1992; Jenik 1994; Bond 2000). Therefore, the intriguing tradeoff between investment on attracting pollinator or in rewarding them could be partly the product of differences among studied trees in age and size level of aggregation. To address this question will be the major aim of this of my proposed research in Doñana.