https://doi.org/10.5061/dryad.905qftts6

Description of the data and file structure

Experimental protocols

Experiments were designed to compare plants on which flowers wilted naturally with plants where some or all flowers were manipulated such that they had minimal or no wilting. However, this was achieved differently for the three experimental years.

Experiments were specifically designed to test Hypotheses 1, 2 and 3 in 2019, 2018 and 2020 respectively.

Hypothesis 1: Re-use of salvaged resources from wilted flowers by the same flowers (2019 experiments; Fig. 2 H1).

The 2019 experiments were designed as follows to test hypothesis 1 that resources, salvaged by a plant during a flowering season when its flowers wilt, are reused by the plant during that season for reproduction by the same flowers.

For one group of plants (i.e., R-plants), petals were removed from all flowers when they reached age 6 day, thus preventing petal wilting. As flowers reached this age, each petal was cut at about 5mm from its base (Fig. 1c) and removed from the flowers, thus preventing possible salvage of resources from these removed petals. Furthermore, nectar production by flowers should not have been affected by this treatment as nectar accumulation is restricted to the lowermost part of the corolla, below where petals were cut, and all nectar is ultimately absorbed by each flower resulting in zero net nectar production.

On other plants (i.e., W-plants), there was no manipulation of flower petals, and thus normal petal wilting could occur. Net nectar production by flowers on these plants should also have been zero, as all nectar is ultimately absorbed by each flower.

For both groups of plants, all flowers on each experimental plant were considered test flowers and so were artificially pollinated each day they were open and collected as fruit 5-6 weeks after experimental treatments were completed. Consequently, individual flowers were typically pollinated multiple times.

Hypothesis 1 predicts that seed set and seed weight per flower and per plant would average greater for the plants with flowers allowed to wilt (i.e., W-plants) than for plants where flower petals were removed and therefore prevented from wilting (i.e., R-plants). It was also hypothesised (Hypothesis 3) that the W-plants would be more likely to flower again during the next season than the R-plants.

Hypothesis 2: Re-use of salvaged resources from wilted flowers in neighbouring flowers on the same plant (2018 experiments; Fig. 2 H2)

The 2018 experiments were designed as follows to test hypothesis 2 that resources, salvaged by a plant during a flowering season when its flowers wilt, are reused by the plant during that season for reproduction by adjacent flowers on the same plant. In this case, the reuse of resources from flower wilting by the same flowers was prevented (see below).

Firstly, the flowers on each plant were divided into lower and upper halves, with the flowers in the lower half of the flowering stem having styles cut when they opened and either having no further treatment (i.e., W plants with wilting flowers) or having flowers removed upon reaching age 6 days (i.e., R plants with flowers removed). Thus, these flowers were prevented from forming fruits and seeds, and were either removed or allowed to wilt. Transfer of salvaged resources from flower wilting to seeds of the same flowers was therefore prevented. The lower half included n/2 flowers if the number n of flowers per stem was even and included (n+1)/2 flowers if n was odd.

Secondly, the flowers in the upper half were designated test flowers, artificially pollinated each day they were open, and collected as fruits after another 5-6 weeks. Transfer of salvaged resources from wilted flowers to seeds of adjacent test flowers was therefore possible.

It was therefore hypothesised, similarly to Hypothesis 1, that seed set and seed weight per flower for the test flowers and per plant would average greater for the plants with flowers allowed to wilt (i.e., W-plants) than for plants where flowers were removed and therefore prevented from wilting (i.e., R-plants). However, in this case, the test flowers were adjacent to the flowers that wilted and not the same flowers. It was also hypothesised (Hypothesis 3) that the W-plants would be more likely to flower again during the next season than the R-plants.

Hypothesis 3: Re-use of salvaged resources from wilted flowers during the next flowering season (All experiments; Fig. 2 H3)

The 2020 experiments were specifically designed as follows to test Hypothesis 3 that resources, salvaged by a plant during a flowering season when its flowers wilt, are reused by the plant for reproduction during the next flowering season by the same plant. In this case, reuse of salvaged resources from flower wilting for seed production by the same plants during the initial season was prevented (see below).

In this case, there were two groups of plants differing only in terms of whether all flowers were allowed to wilt (i.e., W plants) or were prevented from doing so (i.e., R plants), with no difference between these two groups in terms of seed production, as seed production was impossible for both groups. For the R plants, removing each flower when first open (i.e., age day 1) prevented these flowers from wilting, and obviously also prevented seed production by these flowers. For the W plants, when each flower was first open (i.e., age day 1), removing the stigma along with the style tip, prevented seed production for these flowers.

As neither W nor R plants could set seed, any effect of salvaging resources from wilted flowers would necessarily be delayed until a subsequent reproductive episode, and naturally there was no collection of fruits for this 2020 experiment.

However, whether the plants from this experiment re-flowered during the next flowering season was determined by inspecting these plants at about the same time in 2021 (i.e., 14 Jan 2021).

Similarly, whether the plants from the 2018 and 2019 experiments re-flowered during the next flowering seasons were determined by inspecting these plants at about the same times in 2019 and 2020 respectively (i.e., 15-16 Jan 2019 & 19 Jan 2020).

For experiments in all three years, it was hypothesised that W plants would be more likely to re-flower during the next season than R plants.

Data spreadsheets

1.Shadehouse-Flower wilting #11-16 Jan to 31 Jan 2019 plus fruit & seed count data & seed wgt data per flower with pollinations-Dryad

Treatment:       R (petals removed)      W (petals allowed to wilt)

N_Treat:          R = 0                           W = 1

Stem hgt:         Height of flowering stem in cm

Num_Fls:        Number of flowers per stem or plant

Plant:               Plant ID

Flower:            Flower position, numbering from topmost

Seeds_wgt_mg:          Weight of seeds per flower (mg)

N_Seeds:         Number of seeds per flower

Num_Pollinations:      Number of times each flower pollinated

Empty cells are missing data

2.      Shadehouse - Flower wilting #10 - 16 Jan to 31 Jan 2019 plus fruit & seed count, seed wgt & num pollin  data per  plant – Dryad

Plant:               Plant ID

N_Treat:          R (petals removed) = 0                          W (petals allowed to wilt)= 1

Stem hgt:          Height of flowering stem in cm

Num_Fls:         Number of flowers per stem or plant

N_Pollinations: Number of times each plant pollinated

Seeds_wgt per plant:     Weight of seeds per plant (mg)

Num_Seeds per plant:   Number of seeds per plant

Empty cells are missing data

3.      Shadehouse - Flower wilting expt - 15 Jan to 2 Feb 2018 - Seeds per flower & seed wgts with num pollinations – Dryad

Plant_ID:         Plant ID

Treatment:       R (petals removed)      W (petals allowed to wilt)

N_Treat:          R = 0                           W = 1

Stem hgt:         Height of flowering stem in cm

Num_Fls:        Number of flowers per stem or plant

Flower:            Flower position, numbering from topmost

N_Pollinations:           Number of times each flower pollinated

N_Seeds:         Number of seeds per flower

Seeds_wgt_mg:            Weight of seeds per flower

Empty cells are missing data

4.      Shadehouse - Flower wilting expt - 15 Jan to 2 Feb 2018 - Seeds num & wgt etc per plant with num pollinations – Dryad

Plant_ID:         Plant ID

N_Treat:          R (flowers removed)= 0                        W (flowers allowed to wilt)= 1

Stem hgt:          Height of flowering stem in cm

N_Pollinations: Number of times each plant pollinated

Num_Fls:         Number of flowers per stem or plant

Num_Seeds:     Number of seeds per plant

Seeds_wgt_mg:            Weight of seeds per plant

5.      Shadehouse 2019 - 2021  - Flower wilting- whether flowering in year n+1 vs treatment in year n – Dryad

Year-Plant_ID  Plant_ID for each year

Year after treatment      Year following treatment and when flowering recorded

Treatment        R (petals or flowers removed)               W (petals and flowers allowed to wilt)

Whether flowering one year after treatment        1 is yes 0 is no

Empty cells are missing data