DEFENSIVE FUNCTIONS AND POTENTIAL ECOLOGICAL CONFLICTS OF FLORAL STICKINESS

View/Open
This file includes the data for the experiments on the "Effect of stickiness on fruit set" and the "Effects of insects trapped on sticky petals on fruit set": Effect of stickiness on fruit set To test for a causal relationship between stickiness, herbivory, and plant reproduction, we selected two inflorescences on each of ten different plants that produce sticky flowers. Just before anthesis, all of the buds of one of the inflorescences were each washed with 100 μL 95% methanol on a Q-Tip to remove the stickiness (MeOH treatment); the buds on the other inflorescence were kept as a control (Control treatment) and 100 μL of MeOH was similarly applied to the pedicels as a sham control for the MeOH treatment. One week after the MeOH application (flowers in full bloom and just beginning to wilt), we assessed floral damage as the number of damaged flowers relative to the total number of flowers in each inflorescence. Inflorescences contained between 2 and 7 flowers. Fruit set was measured as the number of seed pods per initial number of flowers on each infructescence 100 days after the initial treatment. The data were analyzed by using binomial regression using the relative damage and relative fruit set as the response variable, treatment as independent factor, and the individual plant as a random factor. In addition to the field experiments, we conducted laboratory feeding assays with two generalist herbivore species not commonly found in the plant to test for the actual feeding resistance rather than stickiness effect in the two different plant morphs. Effect of insects trapped on sticky petals on fruit set. To test for a potential proto-carnivorous, indirect defensive, or pollinator-attracting function of stickiness, in the Campo Alegre population, we selected two inflorescences on each of ten different plants producing sticky petals and applied one of two treatments: No insects treatment, insects were removed manually from each bud in the inflorescence without removing the stickiness; Insect treatment, in this treatment the insects removed in the previous treatment were added for an increased number of insects trapped to each bud. We measured relative damage seven days and relative fruit set 100 days after the initial treatment, as described above. The data were analyzed by using binomial regression using the relative damage and relative fruit set as the response variable, treatment as independent factor, and the individual plant as a random factor.  [Microsoft Excel]   (13.76Kb)
This file includes all raw data on the analysis of the non-volatile compounds based on LC-MS analyses: To analyze the non-volatile corolla surface chemistry, petals from both sticky and non-sticky plants were collected from different plants of those used for other experiments and washed with methanol to remove the sticky layer. These extracts were analyzed on an LC Dionex UltiMate 3000 (Thermo Scientific, Germering, Germany) equipped with a degassing unit, a gradient binary pump, and an autosampler with 120-vial well-plate trays, and a thermostatically controlled column compartment. Chromatographic separation was performed on a Hypersil GOLD aQ C18 column (Thermo Scientific, Sunnyvale, CA, USA, 100 mm x 2.1 mm i.d., 1.9 μm particle size). The column temperature was 35ºC, flow rate 0.3 mL/min, injection volume 2.0 μL, the mobile phase was acetonitrile grade LC/MS modified with 0.2% v/v formic acid (A) and an aqueous solution with 0.2% v/v formic acid (B). The initial gradient condition was 100% A, changed linearly to 100% B in 8 min, maintained for 4 min, returned to 100% A in 1 min, and maintained for 3 min. The injection volume was 2 μL. The LC was connected to an Exactive Plus Orbitrap mass spectrometer (Thermo Scientific, Bremen, Germany) with a heated-electrospray ionization (HESI-II) source operated in the positive ion mode. The capillary voltage was set at 3.5 kV. The nebulizer and capillary temperatures were set at 350 and 320 °C, respectively; sheath gas and auxiliary gas (N2) were adjusted to 40 and 10 arbitrary units, respectively. Nitrogen (>99%) was obtained from a generator (NM32LA, Peak Scientific, Scotland, UK). During the full scan MS, the Orbitrap-MS mass-resolution was set at 70000 (full-width-at-half-maximum, at m/z 200, R FWHM ) with automatic gain control (AGC) target, 3x10 6, C-trap maximum injection time, 200 ms, and a scan range of m/z 100–1000. The ions injected into the HCD-cell via the C-trap were fragmented with stepped-normalized collision energies of 10, 20, 30, and 40 eV. The mass spectra were recorded in the AIF (All-ion fragmentation) mode for each collision energy at R FWHM of 35000, AGC target, 3x10 6, C-trap injection time, 50 ms, and a mass range of m/z 80–1000. The data obtained were analyzed using Thermo Xcalibur 3.1 software (Thermo Scientific, San Jose, CA, USA). Compound identification was based on exact mass measurement ([M]+ or [M+H]+), elemental composition calculation for both protonated molecules and their product ions, and the comparison of retention times of reference substances with those in flower extract´s chromatograms. The metabolomics (METLINTM, http://metlin.scripps.edu) and phytochemistry (PCIDB, http://www.genome.jp/db/pcodb) databases were employed. With the relative peak area data, we compared the changes in the compositions between sticky and non-sticky flowers using permutation analysis of variances (PERMANOVA) and a Nonmetric multidimensional scaling (NMDS) on Bray-Curtis distances for visualization. Moreover, we conducted a random forest (RF) analysis using the packages randomForest 59 and varSelRF 60 in R version 3.3.1 58. To do this, we first conducted an RF classification analysis for sticky and non-sticky flowers. We then used 200 bootstrap iterations to select the compounds that best distinguished between sticky and non-sticky, followed by a t-test with the resulting compounds.  [CSV file]   (9.098Kb)
This file includes all raw data on the floral emissions of volatile organic compounds quantified and identified by GC-MS: We collected volatiles following the headspace of flowers (Sticky n=8, non-sticky n=13) and leaves (sticky n=10, non-sticky n=13) from plants by enclosing them in 500 mL polyethylene cups fitted with ORBO-32 charcoal adsorbent tubes (Supelco, Bellefonte, PA, USA). Air was pulled through the cup at a flow rate of approximately 150 mL/min for 8 hours using an active air sampling vacuum pump (IONTIK, USA). Five additional charcoal tubes collected air from the environment as a control. Tubes were then capped and kept frozen before analysis. Before elution, we added 5 µL tetraline (90 ng/mL) as an internal standard to each tube. The tubes were then desorbed with 350 mL of dichloromethane and samples were analyzed in a Varian CP-3800 GC coupled to a Saturn 2200 MS and fit with a DB-WAX column (J&W Scientific) of 60 m × 0.25 mm id capillary column coated with polyethyleneglycol (0.25 mm film thickness). Helium (99.995%, Airgas®) was used as a carrier gas. Total ion chromatograms were integrated, and peak areas of individual compounds were normalized by the area of the internal standard. Tentative identification was made by comparing the mass spectrum with the NIST Mass Spectral Database. As before, we compared the changes in the compositions between sticky and non-sticky flowers and leaves using permutation analysis of variances (PERMANOVA) and a Nonmetric multidimensional scaling (NMDS) on Bray-Curtis distances for visualization.  [CSV file]   (12.79Kb)
This file includes the raw data of a field "Hummingbird exclusion" experiment: Hummingbird exclusion Based on the findings in the first part of this study, we hypothesized that the population-specific differences in the effects of stickiness on florivory could be due to the differences in the presence of hummingbirds. Sticky plants are efficiently excluding interactions with insects both mutualist and antagonist. Thus, in populations with a high density of hummingbirds such as in the CA population, sticky plants will reap the benefits of protecting their flowers from herbivory without compromising pollination by hummingbirds. In such a population, non-sticky plants will have lower levels of fruit set when florivores are present. On the other hand, in populations with low densities of hummingbirds, sticky plants can be predicted to have a lower rate of pollination due to the additional reduction in pollination by insects. In this case, the costs of not attracting insect pollinators associated with stickiness are predicted to be larger than the benefits of protecting the flowers from florivorous. To test this hypothesis, we chose eight sticky plants and eight non-sticky plants in the Tres Viejas population, and two branches per plant were selected. One of those branches was protected by a net with a mesh of 2 cm, supported on a wire structure that separated the net from the flowers. The holes allowed the major insect pollinators, bees, and bumblebees 27, to visit the flowers but excluded the hummingbirds. Finally, we measured the relative fruit set of each branch by counting the number of flowers that became fruits on those branches. The results were analyzed by a binomial regression with fruit set as dependent variable. Stickiness and the presence of the net were considered independent factors. The plant was considered a random factor. Effects of the factors were assessed by using a likelihood-ratio test, followed by a post hoc test using the CLD function from the ‘emmeans’ in R.  [CSV file]   (1.580Kb)
This file includes al raw data on a flower damage survey across three populations: Survey of floral damage and seed set. To determine if the stickiness affects the proportion of damage to flowers by herbivores, we first surveyed sticky and non-sticky plants in each population, then randomly chose and marked one inflorescence per plant to count the number of flowers and buds and the proportion of those with any damage by florivorous species (rate of herbivory). One month later, we recorded the proportion of those flowers that had turned into fruit (fruit set) as a measure of fitness. In total, we recorded data from 50 plants in CH, 38 in TV, and 38 in CA. We found different proportions of sticky and non-sticky plants in the three populations (CH 2:3, TV 5:3, CA 4:1). During the survey, it became evident that much of the damage was due to a Tortricidae (Lepidoptera) larvae eating the anthers, style, and ovules of the flower, usually starting just before the anthesis of the flowers. So, we conducted a second survey to specifically estimate the amount of damage to buds and flowers affected by the caterpillars in sticky and non-sticky plants. Statistical analyses were done using binomial regression with the total number of flowers or buds and the number of them with damage as dependent variable and stickiness and population as independent factors. Plant ID in each population was used as a random factor. Effects of the factors were assessed by using a likelihood-ratio test, followed by a post hoc test using the CLD function from the ‘emmeans’ in R version 4.1.1 58.  [CSV file]   (14.84Kb)
This file includes fruit set data of sticky and non-sticky plants across three populations: Survey of floral damage and seed set. To determine if the stickiness affects the proportion of damage to flowers by herbivores, we first surveyed sticky and non-sticky plants in each population, then randomly chose and marked one inflorescence per plant to count the number of flowers and buds and the proportion of those with any damage by florivorous species (rate of herbivory). One month later, we recorded the proportion of those flowers that had turned into fruit (fruit set) as a measure of fitness. In total, we recorded data from 50 plants in CH, 38 in TV, and 38 in CA. We found different proportions of sticky and non-sticky plants in the three populations (CH 2:3, TV 5:3, CA 4:1). During the survey, it became evident that much of the damage was due to a Tortricidae (Lepidoptera) larvae eating the anthers, style, and ovules of the flower, usually starting just before the anthesis of the flowers. So, we conducted a second survey to specifically estimate the amount of damage to buds and flowers affected by the caterpillars in sticky and non-sticky plants. Statistical analyses were done using binomial regression with the total number of flowers or buds and the number of them with damage as dependent variable and stickiness and population as independent factors. Plant ID in each population was used as a random factor. Effects of the factors were assessed by using a likelihood-ratio test, followed by a post hoc test using the CLD function from the ‘emmeans’ in R version 4.1.1 58.  [CSV file]   (3.968Kb)
PERMANENT LINK(S)
https://hdl.handle.net/1813/111390
Collections
Metadata
Show full item record
Author
Chautá, Alexander; Kumar, Arvind; Mejia, Jessica; Stashenko, Elena; Kessler, Andre
Abstract
Stickiness of vegetative tissues has evolved multiple times in different plant families but is rare and understudied in flowers. While stickiness in general is thought to function primarily as a defense against herbivores, it can compromise mutualistic interactions (such as those with pollinators) in reproductive tissues. Here, we test the hypothesis that stickiness on flower petals of the High-Andean plant, Bejaria resinosa (Ericaceae), functions as a defense against florivores. We address ecological consequences and discuss potential trade-offs associated with a repellant trait expressed in flowers that mediate mutualistic interactions. In surveys and manipulative experiments, we assess florivory and resulting fitness effects on plants with sticky and non-sticky flowers in different native populations of B. resinosa in Colombia. In addition, we analyze the volatile and non-volatile components in sticky and non-sticky flower morphs to understand the chemical information context within which stickiness is expressed. We demonstrate that fruit set is strongly affected by floral stickiness but also varies with population. While identifying floral stickiness as a major defensive function, our data also suggest that the context-dependency of chemical defense functionality likely arises from differential availability of primary pollinators and potential trade-offs between chemical defense with different modes of action.
Description
This dataset includes results from HPLC and GC-MS analyses on the non-volatile and volatile secondary metabolites of B. resinosa. The data are raw signal intensity and signal intensity of each compound relative to an internal standard. The data were the basis for the analyses in the original publication and can be used for subsequent analyses after informing the original authors.
Sponsorship
The research was funded by a grant to AC by Fundación CEIBA (Centro de Estudios Interdisciplinarios Básicos y Aplicados) and a grant from the New Phytologist Foundation to AK.
Date Issued
2022
Publisher
TBD
Subject
ecological conflicts; indirect defenses; plant defense; pollination; protocarnivory; secondary metabolites
Related Version
Raw Data associated with original publication
Previously Published As
TBD
Rights
CC0 1.0 Universal
Rights URI
Type
article
 
dataset
 
Accessibility Feature
alternative text
Accessibility Hazard
none
Accessibility Summary
Data are accessible once the original paper has been published

The following license files are associated with this item:

    Creative CommonsCC0 1.0 Universal
    Except where otherwise noted, this item's license is described as CC0 1.0 Universal

    Statistics