Heterospecific pollen deposition (HPD) is ubiquitous across plant communities, especially for generalized species which use a diversity of pollinators, and may have negative effects on plant reproduction. However, it is unclear whether temporal changes in the co-flowering community result in changes in HPD patterns. Moreover, community-level studies are required to understand which factors influence HPD and how the reproduction of different species is affected. We investigated the temporal variation of HPD, its relationship with level of specialization on pollinators and floral phenotypic specialization, and its association with reproductive success (pollen limitation and fruit set) in 31 hummingbird-pollinated plant species in a tropical Campo Rupestre. We found seasonality in HPD, with species flowering in the dry season having greater diversity of heterospecific pollen on stigmas and a higher frequency of stigmas containing heterospecific pollen, compared to the rainy season. Stigmas of ecologically generalized species had more heterospecific pollen, while the relationship for ecologically specialized species depended on floral phenotype. Surprisingly, and in contrast to theory, we found a positive relationship between HPD and reproductive success. Our results indicate benefits of generalization and facilitation, in which sharing pollinators brings greater reproductive success via increased conspecific pollen deposition, even if it incurs more HPD. We demonstrated how assessing HPD at a community-level can contribute to understanding the ecological causes and functional consequences of pollinator sharing.

Study Area

Data was collected between August 2017 and July 2018 in a 50 ha area of Campo Rupestre on the JK campus of the Federal University of Vales do Jequitinhonha and Mucuri (UFVJM) in Diamantina/ MG (18°11'48.23"S, 43°34'8.74" W). The sampling area is continuous with Biribiri State Park, a permanent preservation area of more than 16,000 ha. The climate is characterized as mesothermal Cwb type, according to Köppen-Geiger classification, being temperate humid with dry winters (April to September) and wet summers (October to March). The average altitude is 1,281 m with an average temperature of 18.8°C and average annual rainfall of 1,498 mm. Campos Rupestres are regions with herbaceous shrub vegetation established in rocky outcrops with heterogeneity that allows the establishment of several plant species with common characteristics among hummingbird-pollinated plants.

Characterization of the plant-hummingbird community

The observations of plant species visited by hummingbirds were made every week over a year in two 500m x 10m transections, totalling one hectare of sampling area. In each transection, nine fixed plots of 2m² were used for the observation sessions. Each observation session lasted 20min. Observations took place twice a week, with a duration of three hours (nine observation sessions), with morning (09:00 – 12:00) and afternoon (13:00 – 16:00) observations, totalling 288 observation hours. For interaction observations, 80x30 binoculars were used.

Ecological and phenotypic specialization

We focused on number of hummingbird partners as a measurement of ecological specialization. This is because interactions were recorded regardless of the flower availability of each plant species in each plot. A sampling effort of 15 hours per species (regardless of flower abundance) was shown to be sufficient to describe plant-hummingbird interactions (Vizentin-Bugoni et al. 2016). Accordingly, sampling effort was approximately 20 hours for each plant species, distributed across the phenology of each species.  We classified each species following the concept of ecological specialization (Ollerton et al. 2007, Ollerton 2021): ecologically specialized (pollinated by one hummingbird species) and ecologically generalized (pollinated by more than one hummingbird species).

The length and width of the opening of the corolla were measured as descriptors of phenotypic specialization. We measured corolla length and opening width from two flowers of five indviduals, totalling 10 flowers per species. To determine the accessibility of nectar to hummingbirds, we measured the effective length of the corolla, that is, from the receptacle to the corolla opening. We defined the corolla opening as the largest width of the floral tube entrance. Using these data we calculated the average corolla length and opening width for each species.

Pollen receipt patterns

Monthly collections of stigmas of the plant species were performed to evaluate conspecific pollen deposition (CPD) and heterospecific pollen deposition (HPD) for 12 months. We sampled HPD and CPD from 31 of the 57 hummingbird-visited species that were flowering during the collection period (we excluded species with less than three flowering individuals). The samples were collected at the end of flower anthesis (signs of wilting) and between 16:00 and 18:00 to ensure maximum pollen deposition. Two flowers were collected from five individuals per species, totalling 6-10 flowers per species per month. We used this sampling scheme in order to include more species per month (several species were represented only by 5-10 flowering individuals per month) and to avoid oversampling abundant species. We averaged samples per individual. The collection of stigmas was distributed along the flowering period of each plant species, sampling 20-116 stigmas per species. Slides of the stigmas were mounted using glycerin gelatin stained with basic fuchsin for better visualization of pollen under optical microscopy. In addition, slides with pollen from anthers of those 31 species were mounted to construct a reference pollen library. Pollen grains were identified and quantified according to morphological characteristics and classified as conspecific and heterospecific. We calculated the mean CPD/HPD intensity (amount of conspecific or heterospecific pollen) and HPD frequency (proportion of stigmas containing heterospecific pollen) per species (lumping all stigmas collected for the whole 12 months). We have not included CPD frequency because most stigmas contained conspecific pollen (96% of stigmas on average per species). Additionally, we also calculated the proportion of HPD relative to the total pollen load (HPD + CPD) per plant species (hereafter ‘HPD proportion’).

Datasets of “Heterospecific pollen deposition is positively associated with reproductive success in a diverse hummingbird-pollinated plant community”

Published in Oikos

DOI: https://doi.org/10.1111/oik.08714

Authors: Sabrina Lopes, Pedro J Bergamo, Steffani Najara, Jeff Ollerton, Thiago Santos, André Rech

Funding: PJB thanks CNPq (152417/2020-6 fellowship) and FAPERJ (E-26/201.867/2020). AR and SL thanks CNPq (Grant Proc. 400904/2019-5 and Scholarship Proc. 5682018 respectively), SN thanks CAPES (Financial code 001), AR thanks FAPEMIG (Proc. RED-00253-16). 

Two datasets are described here:

1) species_data.xlsx contains species-level metrics of pollen deposition, corolla morphology, interaction with hummingbird pollinators and reproductive success

This dataset contains the following variables:

2) HP_temporal.xlsx

This dataset contains the intensity and frequency of heterospecific pollen deposition of each plant species in each month: