Species interactions are one dimension of the niche, and niche overlap arises when two species share an interaction partner. In pollination systems, environmental and biotic factors impact the niche overlap. Here we explored the effects of climate seasonality, plant and bat richness, morphological traits, and phylogenetic distance in shaping the niche overlap of Neotropical bat-plant pollination networks. For that, we used a dataset of 22 bat-plant pollination networks in the Neotropical region. We measured niche overlap in bats and plants with the Morisita-Horn index, ĈH, and then, we used a SAR model to test the relationships between niche overlap and the abiotic and biotic factors. We found a lower niche overlap among bats in communities composed of phylogenetically distant bat species. Moreover, plant and bat overlap were lower in regions with higher precipitation seasonality. Our results indicate that climate seasonality and bat evolutionary history drive niche overlap in Neotropical bat-plant pollination interactions. These findings suggest that a higher precipitation seasonality may promote the emergence of temporal modules reducing niche overlap, probably as a consequence of seasonal species phenologies.  Furthermore, the method used to record the interactions impacts the degree of niche overlap. Interactions recorded with pollen samples tend to have higher niche overlap than direct observations. The uncoupled responses of morphological traits and phylogenetic distances in bat niche overlap suggest an effect of historical processes independently of the morphological traits. Our study reinforced the importance of evolutionary history and ecological processes in imprinting patterns of interaction niche overlap.

1. Bat-plant pollination studies

We obtained published and unpublished lists of pollination interactions between bat and plant species at the community level, performing a search on Google Scholar. For that, we  combined the following keywords: “bat”, “Carolliinae”, “Glossophaginae”, “Lonchophyllinae”, “Phyllostominae”, “Stenodermatinae” with “community”, “flower”, “pollination”, “trophic”, and “chiropterophily”. The results in Google Scholar included published papers, book chapters, theses, dissertations, and undergraduate monographs. Each dataset describes the plant species used as nectar resources by bat species in a given locality. Interactions between bats and plants were summarized in adjacency matrices, with plants in columns and bats in rows. The interaction between a pair of species was represented by 1, while the absence of interaction was 0. The interaction occurrence was described by direct observations (n=5 studies; 23%) or pollen present in the fur or fecal samples of each bat species (n=17 studies; 77%). As some species had outdated taxonomic names, we made taxonomic adjustments. For that, we followed the classifications of Burgin, Colella, Kahn, & Upham (2018) for bats and Flora do Brasil (Jardim Botânico do Rio do Janeiro, 2020) for plants. When a species was not found in Flora do Brasil (2020), we checked in Tropicos v3.2.3 (https://tropicos.org).

 2. Niche overlap

To measure the niche overlap, we used the Morisita-Horn index (ĈH) (Horn, 1966). We quantified niche overlap for bats and plants separately. ĈH describes the mean similarity in interaction patterns between species within the same group (i.e.: plants or bats) (Horn, 1966). To this end, ĈH considers the proportion of species used as a resource by all pairs of species, determining the mean degree of similarity in their resource use within each guild. The index varies from 0 (low niche overlap) to 1 (high niche overlap) (Dormann et al., 2017). We performed ĈH calculations using the function “grouplevel” of the bipartite v2.08 package (Dormann et al., 2017) in R 3.3.3 (R Core Team, 2017).  

3. Drivers of niche overlap

Climate seasonality was described by the standard deviation of monthly temperature and the coefficient of variation of monthly precipitation. We obtained the temperature seasonality (BIO4) and precipitation seasonality (BIO15) from the WorldClim v2.0 database (worldclim.org) at 5 arc-min resolution (Fick and Hijmans, 2017). Most of the studies (n=20; 91%) included in the analyses have a sampling effort equal to or great than 12 months, suggesting that different seasons were recorded in our dataset. 

 We considered bat richness as the number of bat species with nectar consumption records in each community. We included bat richness as a predictor variable in models where we analyzed the drivers of niche overlap of bats. We used the same procedure when analyzing the drivers of the overlap of plants. Hence, we defined plant richness as the number of plant species with records of bat visits in each study. Functional dispersion (FDis) (Laliberté and Legendre, 2010) described the distance in trait filling among species, using only presence-absence data. FDis describes the mean distance of each species to the community’s average trait. Therefore, high values of FDis indicate communities composed of species that are functionally distinct from each other (Laliberté and Legendre, 2010).

We selected traits reflecting morphological adaptations that optimize resource exploitation in bat-plant pollination interactions (Fleming et al., 2009). For bats, we used the forearm length, the greatest length of the skull, the breadth of the braincase, and length of the maxillary tooth row. For plants, we chose traits related to the most common characteristics associated with bat pollination (Fleming et al., 2009). We used plant habit (tree, shrub, herb, liana, or epiphyte), flower accessibility (low, medium, or high), and resource quantity. We classified flower shapes according to their nectar accessibility. Flowers with low accessibility were tubular-shaped; flowers with medium accessibility were  bell-, flag-, funnel-, or cup-shaped, and flowers with high accessibility were open-, dish- or brush-shaped (Ramírez, 2003). Morphological trait characteristics were obtained from the literature. We excluded those species from the analyses because we did not find trait data for 50 plant species (23%). We used the “dbFD” function of the FD v1.0.12 package (Laliberté and Legendre, 2010) in R to quantify functional dispersion.

Phylogenetic distances among species were described by MPD (Webb, 2000). This index represents the mean phylogenetic nodal distance between each species pair in the community while evaluating phylogenetic diversity (Webb, 2000). Therefore, higher values of MPD indicate a community composed of species that are more phylogenetically distant from each other. To calculate MPD, we used the phylogenetic hypothesis of Rojas, Warsi, & Dávalos (2016) for bats and Magallón, Gómez-Acevedo, Sánchez-Reyes, & Hernández-Hernández (2015) for plants. Three bat species (7%) occurring in two communities were missing in the phylogeny: Leptonycteris nivalis, Lonchophylla inexpectata and Xeronycteris vieirai. When the genus was monophyletic, missing species were added to the genus node (L. nivalis and L. inexpectata). When the missing species were monospecific, missing species were added to the subfamily node (Xeronycteris vieirai – Lonchophyllinae, Gardner, 2007).

Species additions were performed in the web service of SUNPLIN (http://bioinfo.inf.ufg.br/sunplin/, Martins, Carmo, Longo, Rosa, & Rangel, 2013). To overcome phylogenetic uncertainty associated with species additions, we built 1,000 phylogenetic hypotheses using the branch-based insertion method (Martins et al., 2013), randomly adding missing species with different branch lengths in specific nodes, as described above (phylogenies available in Appendix S3). We repeated the analyses using all phylogenetic hypotheses to check the robustness of our results. The plant phylogeny was constructed with Phylocom 4.2 (Webb et al., 2008) using the tree of Magallón et al. (2015) as the backbone tree (see Appendix S3, Fig. S3.2). We performed phylogenetic MPD calculations using the “mpd” function of the picante v1.7 package (Kembel et al., 2010) in R. 

References

Burgin, C.J., Colella, J.P., Kahn, P.L., Upham, N.S., 2018. How many species of mammals are there? J. Mammal. 99, 1–14. https://doi.org/10.1093/jmammal/gyx147

Dormann, C.F., Fruend, J., Gruber, B., Devoto, M., Iriondo, J., Strauss, R., Vázquez, D., Blüthgen, N., Clauset, A., Strauss, R., Rodriguez, M., 2017. Package ‘bipartite’: Visualising Bipartite Networks and Calculating Some (Ecological) indices. https://doi.org/10.1002/sim.4177>

Fick, S.E., Hijmans, R.J., 2017. Worldclim 2: New 1-km spatial resolution climate surfaces for global land areas [WWW Document]. Int. J. Climatol. URL http://worldclim.org/version2

Fleming, T.H., Geiselman, C., Kress, W.J., 2009. The evolution of bat pollination: A phylogenetic perspective. Ann. Bot. 104, 1017–1043. https://doi.org/10.1093/aob/mcp197

Gardner, A.L., 2007. Mammals of South America volume 1: Marsupials, Xenarthrans, Shrews and Bats, 1st ed. The University of Chicago Press, Chicago.

Horn, H.S., 1966. Measurement of overlap in comparative ecological studies. Am. Nat. 100, 419–424.

Jardim Botânico do Rio do Janeiro, 2020. Flora do Brasil 2020 [WWW Document]. URL http://floradobrasil.jbrj.gov.br/ (accessed 8.6.21).

Kembel, S.W., Cowan, P.D., Helmus, M.R., Cornell, H. V., Morlon, H., Ackerly, D.D., Blomberg, S.P., Webb, C.O., 2010. Picante: R tools for integrating phylogenies and ecology.

Laliberté, E., Legendre, P., 2010. A distance-based framework for measuring functional diversity from multiple traits A distance-based framework for measuring from multiple traits functional diversity. Ecology 91, 299–305. https://doi.org/10.1890/08-2244.1

Magallón, S., Gómez-Acevedo, S., Sánchez-Reyes, L.L., Hernández-Hernández, T., 2015. A metacalibrated time-tree documents the early rise of flowering plant phylogenetic diversity. New Phytol. 207, 437–453. https://doi.org/10.1111/nph.13264

Martins, W.S., Carmo, W.C., Longo, H.J., Rosa, T.C., Rangel, T.F., 2013. SUNPLIN: Simulation with Uncertainty for Phylogenetic Investigations. BMC Bioinformatics 14. https://doi.org/10.1186/1471-2105-14-324

R Core Team, 2017. R: A language and environment for statistical computing.

Ramírez, N., 2003. Floral specialization and pollination: A quantitative analysis and comparison of the Leppik and the Faegri and van der Pijl classification systems. Taxon 52, 687–700. https://doi.org/10.2307/3647344

Rojas, D., Warsi, O.M., Dávalos, L.M., 2016. Bats (Chiroptera: Noctilionoidea) challenge a recent origin of extant neotropical diversity. Syst. Biol. 65, 432–448. https://doi.org/10.1093/sysbio/syw011

Webb, C.O., 2000. Exploring the phylogenetic structure of ecological communities: An example for rain forest trees. Am. Nat. 156, 145–155. https://doi.org/10.1086/303378

Webb, C.O., Ackerly, D.D., Kembel, S.W., 2008. Phylocom: software for the analysis of phylogenetic community structure and trait evolution. Bioinformatics 24, 2098–2100.

Our data are CSV files. Thus, MS Excel software, Google Browser's extension, OpenOffice, or another software that reads spreadsheets is required. Files can also be opened with R, R Studio, or another statistics software.