Results

At the end of the experiment, we successfully treated 1320 flowers (660 flower pairs). Zero counts made up 24% of the total data, and seed number ranged from 1 to 1282 seeds per flower (depending on species), with a median of 4 seeds per flower.

The effect of pollen type, recipient and donor status on seed set and seed number

Heterospecific pollen did not reduce seed set, and neither seed number (see Figure S 2-3, Table S 3-4. Overall, rare species tended to have higher seed number compared to common species, but this trend was not significant. Similarly, recipient status and donor status did not affect seed set nor seed number (see Figure S 4-5, Table S 5-6).

The effect of recipient and donor self-compatibility on seed set and seed number

Self-incompatible recipient species showed a lower seed set probability compared to self-compatible species (SI/SC odds ratio = 0.0206, SE = 0.0416, p-Value = 0.05). This effect was independent of pollen type treatment (see Figure 2 and Table 1-2). Seed number was not affected by self-compatibility of recipient and donor species). (see Figure S 6-8; Table S 7-8).

The effect of recipient-donor relatedness on seed number

Overall, phylogenetic relatedness between recipient and donor species did not affect the strength of HPI. For rare recipients with rare donors, HPI tended to decrease with relatedness (ΔlrrPD= –0.0017, SE = 0.0010, p-Value = 0.08), with more distantly related recipient-donor species pairs less affected by HP. (see Figure 3 and Table 3-4).