Causes of spatial patterns of fruit set in waratah: Temporal vs. spatial interactions between flowers on an inflorescence

Abstract Spatial patterns of fruit set within inflorescences may be controlled by pollination, nutrient allocation, or inflorescence architecture. Generally, flowers that have spatial and/or temporal precedence are more likely to set fruits. We sought to separate these factors by comparing patterns of fruit set on inflorescences of two species ofTelopea(Proteaceae); one that flowers from the tip to the base of the rachis, the other from base to tip. In both species, most fruits were set at the top of the inflorescence (the last flowers to open forT. speciosissima) and this was extreme forT. mongaensis, where the top flowers open first. Fruit set was not generally limited by inadequate pollination for eitherT. mongaensisorT. speciosissima, as hand pollinations did not increase fruit set and many abscised flowers contained pollen tubes. InT. speciosissima, we tested whether removal of developing topmost fruits would ‘release’ those that had initiated but not yet aborted lower down. There was no significant effect. Plant hormones can increase the degree to which a developing fruit is a sink for nutrients, so we applied cytokinin to the developing lower fruits on some inflorescences. There was no significant effect of the hormone treatment. We conclude that temporal precedence may contribute to the skewed pattern of fruit set inT. mongaensis, because there was an extreme concentration of fruit set on the distal part of the inflorescences, but it cannot explain this pattern of fruit set inT. speciosissima, where the distal flowers are the last to open. Some other process must therefore constrain fruit set to the topmost flowers in an inflorescence. While cytokinin application had no significant effect, the power of this experiment was low and we consider that the hypothesis of hormonal control is worth further exploration.