Post-fire succession niches facilitate co-occurrence of divergent flower forms
Data files
Jul 09, 2026 version files 66.11 KB
-
anthers_removed.csv
2.82 KB
-
density.csv
10.72 KB
-
natural_seed_set.csv
27.54 KB
-
README.md
3.49 KB
-
tube_lengths.csv
20.45 KB
-
visitation.csv
1.09 KB
Abstract
Pollinator-mediated floral divergence is common, but the ecological mechanisms allowing divergent forms to coexist are often unclear. In South Africa, Lapeirousia anceps corolla tube length coevolves with the proboscis lengths of its local pollinators, long-proboscid flies. At most sites, proboscis and corolla length are closely matched, but there are occasional bimodal populations, where short- and long-tubed plants coexist. To understand the selective environment that may lead to bimodality and coexistence, we studied a bimodal population over a 21-year period, during which several fires occurred. We quantified plant density, pollinator visitation, corolla tube length distributions, and seed set in relation to tube length throughout the study, and experimentally assessed the contribution of self-pollination by comparing unmanipulated flowers with flowers from which anthers were removed. Early in succession, pollinators and long-tubed plants were more abundant, and they produced more seeds than shorter-tubed plants. Later in succession, shorter-tubed plants produced relatively more seeds and flowered more frequently. In these plants, self-pollination contributed more to seed set, which explains their success in the absence of pollinators. Patterns of selection on tube length shifted in response to pollinator visitation rates, suggesting that the coexistence of short- and long-tubed plants might be facilitated by differences in post-fire successional niches. This illustrates how temporal heterogeneity caused by fire could promote co-occurrence of divergent forms. If so, ongoing changes in natural fire regimes may affect floral diversity.
Dataset DOI: 10.5061/dryad.dncjsxmfk
Description of the data and file structure
We conducted observations in a Lapeirousia anceps population with bimodal distribution of tube lengths near Mamre, Western Cape, South Africa during late October or early November between the years of 2003 and 2023. Our study area consisted of four sub-populations (sites), all within a few hundred meters of one another but with different fire histories.
Files and variables
File: anthers_removed.csv
Description: In 2018 at site A, we measured tube length of 220 Lapeirousia anceps flowers just before anthesis and then removed the anthers from 75 of the flowers. Later, we counted the number of seeds produced by each flower.
Variables
- plant_ID: each plant was assigned a unique identification number.
- tube: length of the floral tube in mm
- seeds: number of seeds produced
- treatment: AR = anthers removed, NP = natural pollination/unmanipulated
File: density.csv
Description: To estimate the density of Lapeirousia anceps plants for each sampled year and site, we threw a 1 m long stick in the air and then counted all flowering plants within a 1 m2 square (~50cm to each side of the stick) where it landed.
Variables
- year: year in which the density was recorded
- site: A, B, C or D = sub-populations of the population in Mamre with different fire histories
- years_since_fire: time in years since the last fire occurred at the site
- density: number of plants counted in a 1 m2 square
File: tube_lengths.csv
Description: To estimate the abundance of short- and long-tubed plants for each sampled year and site, we measured tube length of one flower per plant for randomly selected Lapeirousia anceps plants using digital callipers.
Variables
- year: year in which the floral tube length was measured
- years_since_fire: time in years since the last fire occurred at the site
- site: A, B, C or D = sub-populations of the population in Mamre with different fire histories
- tube_length: length of the floral tube in mm
File: natural_seed_set.csv
Description: To estimate patterns of selection on floral tube length, we measured the tube lengths of Lapeirousia anceps flowers in anthesis for each sampled year and site. Later, we counted the number of seeds produced by each flower.
Variables
- year: year in which the sample was collected
- years_since_fire: time in years since the last fire occurred at the site
- site: A, B, C or D = sub-populations of the population in Mamre with different fire histories
- tube_length: length of the floral tube in mm
- seeds: number of seeds produced
File: visitation.csv
Description: We observed visitation rates of Moegistorhynchus longirostris to Lapeirousia anceps by observing long-tubed flowers in a patch where all flowers were visible to a single observer for every year and site sampled.
Variables
- year: year in which the observation was made
- years_since_fire: time in years since the last fire occurred at the site
- site: A, B, C or D = sub-populations of the population in Mamre with different fire histories
- visitation_rate: number of pollinator visits per flower per hour
- hours: total number of hours observed
Code/software
The files can be viewed in Excel.