Corrigendum

Abstract

Ecology LettersVolume 25, Issue 10 p. 2340-2344 CorrigendumFree Access Corrigendum This article corrects the following: Global trends in the trophic specialisation of flower-visitor networks are explained by current and historical climate Pedro Luna, Fabricio Villalobos, Federico Escobar, Frederico S. Neves, Wesley Dáttilo, Timothée Poisot, Volume 25Issue 1Ecology Letters pages: 113-124 First Published online: November 10, 2021 First published: 13 September 2022 https://doi.org/10.1111/ele.14105AboutSectionsPDF Comments ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat In ‘Global trends in the trophic specialisation of flower-visitor networks are explained by current and historical climate’ Luna et al. (2022) which was published in Volume 25, issue 1 (January 2022), the author would like to correct the results, figures, tables, conclusions section and supplementary file. Recently, Brimacombe et al. (2022) replicated part of the analyses of our recently published article by Luna et al. (2022) and found, in contrast to our published finding, that sampling intensity and network size were the most important variables for explaining trophic specialization of our compiled flower-visitor networks. Upon re-examining our analyses in detail, we discovered that our published analyses included eight networks that should have been omitted because they did not meet criteria for inclusion in the study. In particular, one was a duplicate, and the other seven should not have been considered in our analyses given that the purpose of our study was to exclude networks focused on just one taxonomic group of floral visitors (i.e., hummingbirds). Thus, we reran our analyses with the corrected dataset. We found that some trophic specialization indexes, statistical values (e.g., model deviance and P-values) and importance of some factors explaining the variation in trophic specialisation changed. This is because trophic specialisation indexes are calculated on individual networks and, therefore, the inclusion of an updated dataset (i.e., trophic specialization indexes and associated environmental data) changed our overall model coefficients and the relation between variables. However, our main conclusion remains qualitatively unchanged. As we wrote in our original publication, “Our findings show that current and historical environmental factors explain the trophic specialisation of flower-visitor networks regardless of the latitudinal zone and elevational location.”. Here, we provide revised results, figures, tables, and conclusions. The updated dataset and the code used to conduct our analyses are available at https://doi.org/10.6084/m9.figshare.19164374.v1. We appreciate all the effort and time spent by Brimacombe et al. (2022), which helped us to identify flaws in our data handling. Revised results In our corrected version, for both tropical and temperate regions, insects remained as the dominant group of floral visitors with Diptera and Hymenoptera being the main taxa along elevations. Our corrected database now comprises 86 updated networks (instead of 87, as one network was duplicated) distributed across the major landmasses of the globe (Figure 1). Figure 1Open in figure viewerPowerPoint Maps showing the location of f lower- visitor networks and their trophic specialisation according to three different indices used in this study. Larger circles indicate higher elevation of the sampled network. Coloured gradients show the values of trophic specialisation of the different indices used in this study; warmer colours denote lower trophic specialisation Comparing our corrected results with the previously reported findings, we observed that mean annual precipitation ceased to be a factor explaining trophic specialisation in the models for i) niche overlap and ii) mean normalised degree, and it was also excluded from the model explaining iii) linkage density due to high co-linearity (Variance Inflation Factor, VIF >3). Moreover, historical temperature stability was excluded from two statistical models (i and ii) due its high co-linearity with other variables (VIF >3). Specifically, our corrected results showed that animals' niche overlap (NO) of flower-visitor networks was not associated to any environmental variable (Table 1). This result differs from our original findings, where mean annual precipitation and historical temperature stability were, respectively, negatively, and positively associated to the animals' niche overlap. As previously reported, variation in animals' NO was not explained by elevation (χ2 = 0.001, df = 1, p = 0.99), latitudinal zone (χ2 = 0.97, df = 1, p = 0.32), or the statistical interaction between these two factors (χ2 = 0.28, df = 1, p = 0.59). Moreover, we found that the values of mean normalised degree (NDmean) decreased with increasing net primary productivity (Table 1, Figure 2a), agreeing with our original results, and indicating that trophic specialization of flower-visitor networks increases with primary productivity. In our original publication, we showed that mean normalised degree was negatively associated with mean annual precipitation and positively associated to historical temperature stability. Instead, our revised analyses now showed that mean normalised degree increase with increasing mean annual temperature (Figure 2b) and historical precipitation stability (Figure 2c). Accordingly, trophic specialisation of flower-visitor networks decreases in sites with higher annual temperatures and precipitation stability. Again, as previously reported, variation in NDmean was not explained by elevation (χ2 = 0.8, df = 1, p = 0.36), latitudinal zone (χ2 = 0.36, df = 1, p = 0.54), or their statistical interaction (χ2 = 0.95, df = 1, p = 0.32). Table 1. Deviance table of the relationships between trophic specialisation and environmental factors Model Environmental factor χ2 df P Animals' niche overlap Net primary productivity 0.1 1 0.75 Mean annual temperature 3.53 1 0.06 Mean annual precipitation 0.76 1 0.38 Historical precipitation stability 1.84 1 0.17 Elevation 0.1 1 0.75 Marginal R2 = 0.08; Conditional R2 = 0.39 Moran's I = −0.01 ± 0.008, P = 0.83 Mean normalized degree Net primary productivity 5.07 1 0.02 Mean annual temperature 13.31 1 0.001 Mean annual precipitation 0.4 1 0.52 Historical precipitation stability 10.77 1 0.001 Elevation 3.07 1 0.079 Marginal R2 = 0.36; Conditional R2 = 0.68 Moran's I = −0.01 ± 0.008, P = 0.65 Linkage density Net primary productivity 4.41 1 0.03 Mean annual temperature 1.57 1 0.2 Historical temperature stability 1.49 1 0.22 Historical precipitation stability 1.23 1 0.26 Elevation 1.77 1 0.18 Marginal R2 = 0.18; Conditional R2 = 0.9 Moran's I = −0.008 ± 0.008, P = 0.71 Figure 2Open in figure viewerPowerPoint Relationship of mean normalised degree with net primary productivity (a), mean annual temperature (b) and historical precipitation stability (c). Relationship of linkage density with net primary productivity (d). The P- values and proportion of explained deviance are shown in Table 1. Note that removal of apparent outliers does not change the results In our original results, we reported that linkage density was not associated with any environmental or spatial factor. However, our revised analyses showed that linkage density (LD) increases with net primary productivity (Figure 2d), indicating that trophic specialisation of flower-visitor networks decreases with productivity. As in our previous findings, variation in LD was also not explained by elevation (χ2 = 2.18, df = 1, p = 0.13), latitudinal zone (χ2 = 1.18, df = 1, p = 0.27), or their statistical interaction (χ2 = 0.72, df = 1, p = 0.39). Finally, our revised results are consistent with our original findings regarding the lack of a spatial pattern of trophic specialization, as latitudinal regions and elevation did not explain the variation of the trophic specialisation of flower-visitor networks. Overall, after correcting the data, our statistical models showed that net primary productivity, mean annual temperature and historical precipitation stability are the factors that explain variations in trophic specialization. According to Brimacombe et al. (2022) trophic specialization indexes have strong relationships with sampling intensity and network size. However, Brimacombe et al. (2022) do not interpret this statistical relationship in explicit biological terms nor mention or consider what the so-called “sampling intensity” metric really measures. For instance, in the formula used to calculate sampling intensity (following Schleuning et al., 2012), only the number of interactions and species recorded within the network are used, without considering any sampling unit (i.e., spatial or temporal) or species abundance estimated independently of those recorded in a network. However, some flower-visitor networks may be inherently less connected or have skewed distributions of interactions per species. Thus, the use of the sampling intensity metric may be conflated as “undersampling” by the measure used by Brimacombe et al. (2022). As such, this sampling intensity index quantifies how many interaction events are recorded per species combination, assuming that species-rich sites should have more observations compared to species-poor sites but differs from the conventional sampling effort/intensity measures commonly used in diversity studies by not considering an explicit sampling unit nor abundance differences among species. Additionally, because network size (i.e., the product of the number of plant and floral visitor species) depends on the number of species recorded, we can expect that it could be related to environmental variables that can explain species richness. For example, net primary productivity and temperature are often found as important drivers of species richness over latitudinal and elevational gradients (e.g., Brown, 2014, Peters et al. 2016). Moreover, species richness by itself is not a factor that can explain the mechanisms behind the specialization of biotic interactions (Sexton et al. 2017; Poisot, 2020). In this sense, we also need to consider that species richness and speciation rates can be a result of specialization and not vice versa (Sexton et al. 2017; Poisot, 2020). Because of these properties, we decided not to include network size and sampling intensity as predictor variables in our models. Regardless, both the findings by Brimacombe et al. (2022) and our corrected results highlight the need for developing effective tools to measure sampling adequacy in ecological networks (e.g., interaction accumulation curves, Jordano et al. 2016) and to investigate how large-scale patterns of species interactions may be shaped by the natural history of such interactions at small spatial scales. Revised conclusions As in our original publication, our revised findings showed that both current and historical environmental factors explain the variation in trophic specialization of flower-visitor networks regardless of the latitudinal zone and elevational location. In general, our corrected results showed that floral visitors tend to have more specialized interactions with plants (i.e, lower mean normalised degree) in more productive sites and less specialized interactions (i.e, higher mean normalised degree) in warmer climates. Moreover, we observed that floral visitors have less specialized interactions (i.e, higher mean normalised degree) in sites with higher historical precipitation stability. When interaction frequencies are considered, our revised results also showed that trophic specialization is lower (i.e., higher linkage density) in more productive sites. These results are consistent with our predictions regarding mean annual temperature and net primary productivity, but not with those regarding historical climate stability, latitudinal region, and elevation. In our original discussion, we considered that mean annual temperature did not explain the trophic specialisation of flower-visitor networks. But, our corrected results indicated that mean normalised degree does increase with increasing mean annual temperature. Despite mean annual temperature being usually related to latitude and species richness (Brown et al. 2014), here we did not find any relationship between latitude and trophic specialization. Still, spatial variation of trophic specialization does not need to be related to latitude per se, which does not have biological meaning, but to the variation in resource heterogeneity and availability that may in turn be related to space (MacArthur and Pianka, 1966). As we initially expected, the role of mean annual temperature in determining lower trophic specialization could also be related to the higher functional redundancy that can be found in species-rich sites (Tilman et al. 1997). Indeed, we confirm the previous trends found by Schleuning et al. (2012) where they found that plant richness explains trophic specialization of mutualistic ecological networks. However, caution is needed since temperature can also influence the cost of foraging flights and influence how floral visitors interact over spatial gradients independently of species richness (Petanidou et al., 2018; Classen et al. 2020). In contrast to our previous findings, our new results showed that mean annual precipitation does not influence the number of species with which floral visitors interact. This indicates that precipitation might not be a factor affecting insects foraging and consequently their specialization, as already shown for hummingbird trophic specialization (Dalsgaard et al., 2011). In the case of historical climatic stability, its importance was confirmed in our revised results, but we clarify that lower trophic specialisation is found in sites where rainfall has historically been the prevailing condition. In this case, we postulate that constant water availability is related to lower trophic specialization possibly because generalized species have prevailed in stable environments (Burin et al. 2021) increasing community functional redundancy and therefore reducing its specialization. Moreover, our new results indicate that net primary productivity not only influences the number of species with which floral visitors interact (i.e., mean normalised degree), but also their number of interactions (i.e., linkage density). The positive association between productivity and the number of interactions could be mediated through nectar availability across space (Hawkins et al. 2018), thus influencing how many times each species is visited by the animals. This finding highlights the role of resource availability affecting how floral visitors interact, supporting our main hypothesis on how and why species interactions vary over space (i.e., resource distribution and availability determines trophic niche breadth; MacArthur and Pianka, 1966). Finally, our findings still show that current and historical environmental factors explain the trophic specialisation of flower-visitor networks regardless of the latitudinal zone and elevational location, but this can depend on how trophic specialization is measured. Supporting Information Filename Description ele14105-sup-0001-Supinfo.docWord document, 1.7 MB Appendix S1 Supporting Information Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. REFERENCES Brimacombe, C., Bodner, K. & Fortin, M.J. (2022) How network size strongly determines trophic specialization: a technical comment on Luna et al. (2022). Ecology Letters, 25, 1914– 1916. Wiley Online LibraryPubMedWeb of Science®Google Scholar Brown, J.H. (2014) Why are there so many species in the tropics? Journal of Biogeography, 41, 8– 22. Wiley Online LibraryPubMedWeb of Science®Google Scholar Burin, G., Guimarães, P.R. & Quental, T.B. (2021) Macroevolutionary stability predicts interaction patterns of species in seed dispersal networks. Science (80- ), 372, 733– 737. CrossrefCASPubMedWeb of Science®Google Scholar Classen, A., Eardley, C.D., Hemp, A., Peters, M.K., Peters, R.S., Ssymank, A. et al. (2020) Specialization of plant–pollinator interactions increases with temperature at Mt. Kilimanjaro. Ecology and Evolution, 10(4), 1– 14. Wiley Online LibraryWeb of Science®Google Scholar Dalsgaard, B.O., Magård, E., Fjeldså, J., Martín González, A.M., Rahbek, C., Olesen, J.M. et al. (2011) Specialization in plant- hummingbird networks is associated with species richness, con- temporary precipitation and quaternary climate- change velocity. PLoS One, 6, e25891. CrossrefCASPubMedWeb of Science®Google Scholar Hawkins, B.A., Thomson, J.R. & Mac Nally, R. (2018) Regional patterns of nectar availability in subtropical eastern Australia. Landscape Ecology, 33, 999– 1012. CrossrefWeb of Science®Google Scholar Jordano, P. (2016) Sampling networks of ecological interactions. Functional Ecology, 30, 1883– 1893. Wiley Online LibraryWeb of Science®Google Scholar Luna, P., Villalobos, F., Escobar, F., Neves, F.S. & Dáttilo, W. (2022) Global trends in the trophic specialisation of flower-visitor networks are explained by current and historical climate. Ecology Letters, 25(1), 113– 124. Wiley Online LibraryPubMedWeb of Science®Google Scholar Peters, M.K., Hemp, A., Appelhans, T., Behler, C., Classen, A., Detsch, F. et al. (2016) Predictors of elevational biodiversity gradients change from single taxa to the multi t axa community level. Nature Communications, 7(1), 1– 11. https://doi.org/10.1038/ncomms13736 CrossrefWeb of Science®Google Scholar Petanidou, T., Kallimanis, A.S., Lazarina, M., Tscheulin, T., Devalez, J., Stefanaki, A. et al. (2018) Climate drives plant–pollinator interactions even along small scale climate gradients: the case of the Aegean. Plant Biology, 20, 176– 183. Wiley Online LibraryPubMedWeb of Science®Google Scholar Poisot, T. (2020) The theory of ecological specialization. In: S.M. Scheiner & D.P. Mindell (Eds.) Theory of evolution: principles, concepts, and assumptions. Chicago: University of Chicago Press, p. 232. Google Scholar MacArthur, R.H. & Pianka, E.R. (1966) On optimal use of a patchy environment. American Naturalist, 100, 603– 609. CrossrefWeb of Science®Google Scholar Tilman, D., Knops, J., Wedin, D., Reich, P., Ritchie, M. & Siemann, E. (1997) The influence of functional diversity and composition on ecosystem processes. Science, 277(5330), 1300– 1302. CrossrefCASWeb of Science®Google Scholar Schleuning, M., Fründ, J., Klein, A.-M., Abrahamczyk, S., Alarcón, R., Albrecht, M. et al. (2012) Specialization of mutualistic interaction networks decreases toward tropical latitudes. Current Biology, 22, 1925– 1931. CrossrefCASPubMedWeb of Science®Google Scholar Sexton, J.P., Montiel, J., Shay, J.E., Stephens, M.R. & Slatyer, R.A. (2017) Evolution of ecological niche breadth. Annual Review of Ecology Evolution and Systematics, 13(39), 183– 206. CrossrefWeb of Science®Google Scholar Comments Please enable JavaScript to view the comments powered by Disqus. Volume25, Issue10October 2022Pages 2340-2344 FiguresReferencesRelatedInformation