Ever heard of the Münch pressure flow hypothesis? You have now.

In order to develop, flowers require sugars delivered by the phloem. The delivery plays a crucial role because if it fails, development may cease. Failure is especially prevalent during water limitations and diminishes the number or size of fruits which, in grain crops such as maize, decreases productivity. The failure in maize involves mostly female florets because cross-pollination indicates the pollen is often viable when the female floret is not.

The female florets of maize are convenient experimental tools because they are anatomically simple. The pedicel at the base of each floret has a well-defined, cup-shaped phloem terminus resembling a golf tee holding a round nucellus. At the time of pollination but before fertilization, the florets do not have an embryo and the tissues are entirely maternal, simplifying the genetics. Because of this simplicity, maize florets were used in a new paper recently published in Annals of Botany. The results show that, despite the absence of an embryo, an apoplast is part of the path delivering phloem contents to the female floret of maize. Along the phloem axis, the phloem contents are retained but at the termini they are released to the apoplast non-selectively. The release is controlled by plasma membranes of the parenchyma around the phloem termini in the pedicel. The non-selectivity of the release may be the reason pressure is low in the sieve cells at the termini, driving bulk flow toward the sink. The system involves sucrose hydrolysis in the apoplast, causing glucose to accumulate there. The nucellus absorbs the released compounds selectively and supplies them to the embryo sac and later to an embryo.

Differences in membrane selectivity drive phloem transport to the apoplast from which maize florets develop. Annals of Botany 111(4): 551-562
Abstract
Floral development depends on photosynthetic products delivered by the phloem. Previous work suggested the path to the flower involved either the apoplast or the symplast. The objective of the present work was to determine the path and its mechanism of operation. Maize (Zea mays) plants were grown until pollination. For simplicity, florets were harvested before fertilization to ensure that all tissues were of maternal origin. Because sucrose from phloem is hydrolysed to glucose on its way to the floret, the tissues were imaged and analysed for glucose using an enzyme-based assay. Also, carboxyfluorescein diacetate was fed to the stems and similarly imaged and analysed. The images of live sections revealed that phloem contents were released to the pedicel apoplast below the nucellus of the florets. Glucose or carboxyfluorescein were detected and could be washed out. For carboxyfluorescein, the plasma membranes of the phloem parenchyma appeared to control the release. After release, the nucellus absorbed apoplast glucose selectively, rejecting carboxyfluorescein. Despite the absence of an embryo, the apoplast below the nucellus was a depot for phloem contents, and the strictly symplast path is rejected. Because glucose and carboxyfluorescein were released non-selectively, the path to the floret resembled the one later when an embryo is present. The non-selective release indicates that turgor at phloem termini cannot balance the full osmotic potential of the phloem contents and would create a downward pressure gradient driving bulk flow toward the sink. Such a gradient was previously measured by Fisher and Cash-Clark in wheat. At the same time, selective absorption from the apoplast by the nucellar membranes would support full turgor in this tissue, isolating the embryo sac from the maternal plant. The isolation should continue later when an embryo develops.