Trees realize multi-objective optimization to fulfill growth and reproduction, resulting in a definite architecture at a time. Tree structure information is increasingly available via remote sensing and databases, but deciphering such traits multifunctionality and structure-environment interactions remain major challenges, hindering meaningful contributions of such data to plant performance – environment and climate effects studies. Here we investigate how tree functional-structural trade-offs, sharpened by harsh subalpine climate, are modulated by local canopy openness. We ask if a mixed life-history strategy tree species may exhibit more than a single reproductive optimum, represented by a long-term fruit yield, depending on such structure-environment interactions.  The single-image photogrammetry and a semi-automated crown delineation methods were used to obtain a comprehensive set of architectural variables, quantifying crown size, proportions, and metabolic costs for 131 individual rowan trees, growing on the northern slope of the Babia Góra massif in the West Carpathian mountains in Poland.We employed the principal component analysis (PCA), following similar studies, to identify the major dimensions of the absolute (size, PCAabs) and the relative traits (PCArel), separately. We then used a novel interaction analysis R tool to reveal the most influential architecture-environment interactions shaping the reproductive performance of rowans.The reproductive architectural strategies of the studied subalpine rowans differed distinctly between the sheltered and the more exposed individuals. Under higher canopy openness, increased fruit yield was achieved by asymmetrically increasing crown horizontal spread. The pattern was opposite in wind-sheltered trees, and in both cases increasing relative crown length strongly facilitated fruit production, noticeably, after reaching a threshold around the median value (0.73 m m-1). The trees were generally stout, and slenderness strongly limited fruiting in more open situations. The structural diversity of rowans was driven by the size of the more reproductive crowns (61.7% PCAabs variance explained) and by the crown proportions of the less reproductive trees (38.5% PCArel variance explained). As a result, we found diverging size, but converging architecture in the more reproductive rowan trees, still with a considerable variation in the relative horizontal vs. vertical allocations (under higher and lower canopy openness, respectively). High metabolic costs of both crown dimensions were rare, supporting our size-limitation hypothesis in the harsh, light-limited conditions....