Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Plasticity and not adaptation is the primary source of temperature-mediated variation in flowering phenology in North America

Nature Ecology & Evolution volume 8, pages 467–476 (2024)Cite this article

Subjects

Abstract

Phenology varies widely over space and time because of its sensitivity to climate. However, whether phenological variation is primarily generated by rapid organismal responses (plasticity) or local adaptation remains unresolved. Here we used 1,038,027 herbarium specimens representing 1,605 species from the continental United States to measure flowering-time sensitivity to temperature over time (Stime) and space (Sspace). By comparing these estimates, we inferred how adaptation and plasticity historically influenced phenology along temperature gradients and how their contributions vary among species with different phenology and native climates and among ecoregions differing in species composition. Parameters Sspace and Stime were positively correlated (r = 0.87), of similar magnitude and more frequently consistent with plasticity than adaptation. Apparent plasticity and adaptation generated earlier flowering in spring, limited responsiveness in late summer and delayed flowering in autumn in response to temperature increases. Nonetheless, ecoregions differed in the relative contributions of adaptation and plasticity, from consistently greater importance of plasticity (for example, southeastern United States plains) to their nearly equal importance throughout the season (for example, Western Sierra Madre Piedmont). Our results support the hypothesis that plasticity is the primary driver of flowering-time variation along temperature gradients, with local adaptation having a widespread but comparatively limited role.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Spatial and temporal relationships between flowering time and temperature resulting from plasticity and adaptation.
Fig. 2: Distributions of and relationship between Sspace and Stime among 1,605 North American angiosperms.
Fig. 3: Variation in apparent plasticity (Stime) and apparent adaptation (Sspace − Stime) attributable to differences in phenological niche and native climate among species.
Fig. 4: Variation in apparent plasticity and apparent adaptation among species with varying phenological niches across ecoregions of the United States.

Similar content being viewed by others

Data availability

The data used in this study are publicly available on Zenodo61.

Code availability

All code necessary to reproduce the main results, Extended Data Figures and Supplementary Information are available on Zenodo61.

References

  1. Elzinga, J. A. et al. Time after time: flowering phenology and biotic interactions. Trends Ecol. Evol. 22, 432–439 (2007).

    PubMed  Google Scholar 

  2. Bradshaw, A. D. in Advances in Genetics Vol. 13 (eds Caspari, E. W. & Thoday, J. M.) 115–155 (Academic, 1965).

  3. Gienapp, P., Teplitsky, C., Alho, J. S., Mills, J. A. & Merilä, J. Climate change and evolution: disentangling environmental and genetic responses. Mol. Ecol. 17, 167–178 (2008).

    CAS  PubMed  Google Scholar 

  4. Hoffmann, A. A. & Sgrò, C. M. Climate change and evolutionary adaptation. Nature 470, 479–485 (2011).

    ADS  CAS  PubMed  Google Scholar 

  5. Franks, S. J., Sim, S. & Weis, A. E. Rapid evolution of flowering time by an annual plant in response to a climate fluctuation. Proc. Natl Acad. Sci. USA 104, 1278–1282 (2007).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  6. Wu, Y. & Colautti, R. I. Evidence for continent-wide convergent evolution and stasis throughout 150 y of a biological invasion. Proc. Natl Acad. Sci. USA 119, e2107584119 (2022).

    CAS  PubMed  PubMed Central  Google Scholar 

  7. Merilä, J. & Hendry, A. P. Climate change, adaptation and phenotypic plasticity: the problem and the evidence. Evol. Appl. 7, 1–14 (2014).

    PubMed  PubMed Central  Google Scholar 

  8. Fox, R. J., Donelson, J. M., Schunter, C., Ravasi, T. & Gaitán-Espitia, J. D. Beyond buying time: the role of plasticity in phenotypic adaptation to rapid environmental change. Philos. Trans. R. Soc. B 374, 20180174 (2019).

    Google Scholar 

  9. Phillimore, A. B., Hadfield, J. D., Jones, O. R. & Smithers, R. J. Differences in spawning date between populations of common frog reveal local adaptation. Proc. Natl Acad. Sci. USA 107, 8292–8297 (2010).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  10. Bonamour, S., Chevin, L.-M., Charmantier, A. & Teplitsky, C. Phenotypic plasticity in response to climate change: the importance of cue variation. Philos. Trans. R. Soc. B 374, 20180178 (2019).

    Google Scholar 

  11. Ensing, D. J. & Eckert, C. G. Interannual variation in season length is linked to strong co-gradient plasticity of phenology in a montane annual plant. New Phytol. 224, 1184–1200 (2019).

    PubMed  Google Scholar 

  12. Stinchcombe, J. R. et al. A latitudinal cline in flowering time in Arabidopsis thaliana modulated by the flowering time gene FRIGIDA. Proc. Natl Acad. Sci. USA 101, 4712–4717 (2004).

    ADS  CAS  PubMed  PubMed Central  Google Scholar 

  13. Montague, J. L., Barrett, S. C. H. & Eckert, C. G. Re-establishment of clinal variation in flowering time among introduced populations of purple loosestrife (Lythrum salicaria, Lythraceae). J. Evol. Biol. 21, 234–245 (2008).

    CAS  PubMed  Google Scholar 

  14. Anderson, J. T., Inouye, D. W., McKinney, A. M., Colautti, R. I. & Mitchell-Olds, T. Phenotypic plasticity and adaptive evolution contribute to advancing flowering phenology in response to climate change. Proc. R. Soc. B 279, 3843–3852 (2012).

    PubMed  PubMed Central  Google Scholar 

  15. Conover, D. O. & Schultz, E. T. Phenotypic similarity and the evolutionary significance of countergradient variation. Trends Ecol. Evol. 10, 248–252 (1995).

    CAS  PubMed  Google Scholar 

  16. Nylin, S. & Gotthard, K. Plasticity in life-history traits. Annu. Rev. Entomol. 43, 63–83 (1998).

    CAS  PubMed  Google Scholar 

  17. Fitter, A. H. & Fitter, R. S. R. Rapid changes in flowering time in British plants. Science 296, 1689–1691 (2002).

    ADS  CAS  PubMed  Google Scholar 

  18. Cook, B. I. et al. Sensitivity of spring phenology to warming across temporal and spatial climate gradients in two independent databases. Ecosystems 15, 1283–1294 (2012).

    Google Scholar 

  19. Lapenis, A., Henry, H., Vuille, M. & Mower, J. Climatic factors controlling plant sensitivity to warming. Clim. Change 122, 723–734 (2014).

    ADS  Google Scholar 

  20. Zhang, H., Yuan, W., Liu, S., Dong, W. & Fu, Y. Sensitivity of flowering phenology to changing temperature in China. J. Geophys. Res. Biogeosci. 120, 1658–1665 (2015).

    Google Scholar 

  21. Park, D. S. et al. Herbarium specimens reveal substantial and unexpected variation in phenological sensitivity across the eastern United States. Philos. Trans. R. Soc. B 374, 20170394 (2019).

    Google Scholar 

  22. Prevéy, J. S. et al. Warming shortens flowering seasons of tundra plant communities. Nat. Ecol. Evol. 3, 45–52 (2019).

    PubMed  Google Scholar 

  23. Delgado, M. D. M. et al. Differences in spatial versus temporal reaction norms for spring and autumn phenological events. Proc. Natl Acad. Sci. USA 117, 31249–31258 (2020).

    ADS  PubMed  PubMed Central  Google Scholar 

  24. Li, D. et al. Climate, urbanization and species traits interactively drive flowering duration. Glob. Change Biol. 27, 892–903 (2021).

    ADS  CAS  Google Scholar 

  25. Davis, C. C., Willis, C. G., Connolly, B., Kelly, C. & Ellison, A. M. Herbarium records are reliable sources of phenological change driven by climate and provide novel insights into species’ phenological cueing mechanisms. Am. J. Bot. 102, 1599–1609 (2015).

    PubMed  Google Scholar 

  26. Willis, C. G. et al. Old plants, new tricks: phenological research using herbarium specimens. Trends Ecol. Evol. 32, 531–546 (2017).

    PubMed  Google Scholar 

  27. Park, D. S., Breckheimer, I. K., Ellison, A. M., Lyra, G. M. & Davis, C. C. Phenological displacement is uncommon among sympatric angiosperms. New Phytol. 233, 1466–1478 (2022).

    PubMed  Google Scholar 

  28. Ramirez-Parada, T. H., Park, I. W. & Mazer, S. J. Herbarium specimens provide reliable estimates of phenological responses to climate at unparalleled taxonomic and spatiotemporal scales. Ecography 2022, e06173 (2022).

    ADS  Google Scholar 

  29. Park, I. W. & Mazer, S. J. Overlooked climate parameters best predict flowering onset: assessing phenological models using the elastic net. Glob. Change Biol. 24, 5972–5984 (2018).

    ADS  Google Scholar 

  30. Park, I. W., Ramirez-Parada, T. & Mazer, S. J. Advancing frost dates have reduced frost risk among most North American angiosperms since 1980. Glob. Change Biol. 27, 165–176 (2021).

    ADS  CAS  Google Scholar 

  31. Kharouba, H. M. & Vellend, M. Flowering time of butterfly nectar food plants is more sensitive to temperature than the timing of butterfly adult flight. J. Anim. Ecol. 84, 1311–1321 (2015).

    PubMed  Google Scholar 

  32. Munson, S. M. & Long, A. L. Climate drives shifts in grass reproductive phenology across the western USA. New Phytol. 213, 1945–1955 (2017).

    PubMed  Google Scholar 

  33. Kopp, C. W., Neto-Bradley, B. M., Lipsen, L. P. J., Sandhar, J. & Smith, S. Herbarium records indicate variation in bloom-time sensitivity to temperature across a geographically diverse region. Int. J. Biometeorol. 64, 873–880 (2020).

  34. Pearson, K. D., Love, N. L. R., Ramirez-Parada, T., Mazer, S. J. & Yost, J. M. Phenological trends in the California poppy (Eschscholzia californica): digitized specimens reveal intraspecific variation in the sensitivity of flowering date to climate change. Madroño 68, 343–359 (2021).

    Google Scholar 

  35. Mazer, S. J., Love, N. L. R., Park, I. W., Ramirez-Parada, T. & Matthews, E. R. Phenological sensitivities in two Clarkia congeners: indirect evidence for facilitation, convergence, niche conservatism, or genetic constraints. Madroño 68, 388–405 (2021).

    Google Scholar 

  36. Park, D. S., Xie, Y., Ellison, A. M., Lyra, G. M. & Davis, C. C. Complex climate-mediated effects of urbanization on plant reproductive phenology and frost risk. New Phytol. 239, 2153–2165 (2023).

    PubMed  Google Scholar 

  37. Kramer, K. Phenotypic plasticity of the phenology of seven European tree species in relation to climatic warming. Plant Cell Environ. 18, 93–104 (1995).

    Google Scholar 

  38. Levin, D. A. Flowering-time plasticity facilitates niche shifts in adjacent populations. New Phytol. 183, 661–666 (2009).

    PubMed  Google Scholar 

  39. De Frenne, P. et al. Temperature effects on forest herbs assessed by warming and transplant experiments along a latitudinal gradient. Glob. Change Biol. 17, 3240–3253 (2011).

    ADS  Google Scholar 

  40. Franks, S. J., Weber, J. J. & Aitken, S. N. Evolutionary and plastic responses to climate change in terrestrial plant populations. Evol. Appl. 7, 123–139 (2014).

    PubMed  Google Scholar 

  41. Vitasse, Y. et al. Elevational adaptation and plasticity in seedling phenology of temperate deciduous tree species. Oecologia 171, 663–678 (2013).

    ADS  PubMed  Google Scholar 

  42. Colautti, R. I. & Barrett, S. C. H. Rapid adaptation to climate facilitates range expansion of an invasive plant. Science 342, 364–366 (2013).

    ADS  CAS  PubMed  Google Scholar 

  43. Ackerly, D. D. et al. Topoclimates, refugia and biotic responses to climate change. Front. Ecol. Environ. 18, 288–297 (2020).

    Google Scholar 

  44. Wolkovich, E. M., Cook, B. I. & Davies, T. J. Progress towards an interdisciplinary science of plant phenology: building predictions across space, time and species diversity. New Phytol. 201, 1156–1162 (2014).

    PubMed  Google Scholar 

  45. Chown, S. et al. Adapting to climate change: a perspective from evolutionary physiology. Clim. Res. 43, 3–15 (2010).

    Google Scholar 

  46. Fu, Y. H. et al. Declining global warming effects on the phenology of spring leaf unfolding. Nature 526, 104–107 (2015).

    ADS  CAS  PubMed  Google Scholar 

  47. Güsewell, S., Furrer, R., Gehrig, R. & Pietragalla, B. Changes in temperature sensitivity of spring phenology with recent climate warming in Switzerland are related to shifts of the preseason. Glob. Change Biol. 23, 5189–5202 (2017).

    ADS  Google Scholar 

  48. Wolkovich, E. M., Chamberlain, C. J., Buonaiuto, D. M., Ettinger, A. K. & Morales-Castilla, I. Integrating experiments to predict interactive cue effects on spring phenology with warming. New Phytol. 235, 1719–1728 (2022).

    CAS  PubMed  Google Scholar 

  49. Iler, A. M., CaraDonna, P. J., Forrest, J. R. K. & Post, E. Demographic consequences of phenological shifts in response to climate change. Annu. Rev. Ecol. Evol. Syst. 52, 221–245 (2021).

  50. De Lisle, S. P., Mäenpää, M. I. & Svensson, E. I. Phenotypic plasticity is aligned with phenological adaptation on both micro- and macroevolutionary timescales. Ecol. Lett. 25, 790–801 (2022).

    PubMed  Google Scholar 

  51. Amasino, R. Seasonal and developmental timing of flowering. Plant J. 61, 1001–1013 (2010).

    CAS  PubMed  Google Scholar 

  52. Wolkovich, E. M. et al. Warming experiments underpredict plant phenological responses to climate change. Nature 485, 494–497 (2012).

    ADS  CAS  PubMed  Google Scholar 

  53. Mazer, S. J. et al. Flowering date of taxonomic families predicts phenological sensitivity to temperature: implications for forecasting the effects of climate change on unstudied taxa. Am. J. Bot. 100, 1381–1397 (2013).

    PubMed  Google Scholar 

  54. Beil, I., Kreyling, J., Meyer, C., Lemcke, N. & Malyshev, A. V. Late to bed, late to rise—warmer autumn temperatures delay spring phenology by delaying dormancy. Glob. Change Biol. 27, 5806–5817 (2021).

    CAS  Google Scholar 

  55. Zhou, Z. et al. Lengthened flowering season under climate warming: evidence from manipulative experiments. Agric. For. Meteorol. 312, 108713 (2022).

    Google Scholar 

  56. Roslin, T. et al. Phenological shifts of abiotic events, producers and consumers across a continent. Nat. Clim. Change https://doi.org/10.1038/s41558-020-00967-7 (2021).

  57. Sherry, R. A. et al. Divergence of reproductive phenology under climate warming. Proc. Natl Acad. Sci. USA 104, 198–202 (2007).

    ADS  CAS  PubMed  Google Scholar 

  58. Chen, L. et al. Leaf senescence exhibits stronger climatic responses during warm than during cold autumns. Nat. Clim. Change 10, 777–780 (2020).

    ADS  CAS  Google Scholar 

  59. Zohner, C. M. et al. Effect of climate warming on the timing of autumn leaf senescence reverses after the summer solstice. Science 381, eadf5098 (2023).

    CAS  PubMed  Google Scholar 

  60. Boyle, B. et al. The taxonomic name resolution service: an online tool for automated standardization of plant names. BMC Bioinforma. 14, 16 (2013).

    Google Scholar 

  61. Ramirez-Parada, T. H. et al. Data and code for: plasticity and not adaptation is the primary source of temperature-mediated variation in flowering phenology in North America (Version 1) [Data set]. Zenodo https://doi.org/10.5281/zenodo.8310387 (2023).

  62. Calinger, K. M., Queenborough, S. & Curtis, P. S. Herbarium specimens reveal the footprint of climate change on flowering trends across north-central North America. Ecol. Lett. 16, 1037–1044 (2013).

    PubMed  PubMed Central  Google Scholar 

  63. Carpenter, B. et al. Stan: a probabilistic programming language. J. Stat. Softw. 76, 1–32 (2017).

  64. Iler, A. M., Inouye, D. W., Schmidt, N. M. & Høye, T. T. Detrending phenological time series improves climate–phenology analyses and reveals evidence of plasticity. Ecology 98, 647–655 (2017).

    PubMed  Google Scholar 

  65. Daru, B. H. et al. Widespread sampling biases in herbaria revealed from large-scale digitization. New Phytol. 217, 939–955 (2018).

    PubMed  Google Scholar 

  66. The PLANTS Database. USDA http://plants.usda.gov (accessed 23 August 2023).

  67. Omernik, J. M. Ecoregions of the conterminous United States. Ann. Assoc. Am. Geogr. 77, 118–125 (1987).

    Google Scholar 

  68. Omernik, J. M. & Griffith, G. E. Ecoregions of the conterminous United States: evolution of a hierarchical spatial framework. Environ. Manag. 54, 1249–1266 (2014).

    ADS  Google Scholar 

  69. Wood, S. N. Generalized Additive Models: An Introduction with R (CRC Press/Taylor & Francis, 2017).

  70. R Core Team. R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2013).

  71. Jin, Y. & Qian, H. V. PhyloMaker: an R package that can generate very large phylogenies for vascular plants. Ecography 42, 1353–1359 (2019).

    ADS  Google Scholar 

  72. Paradis, E., Claude, J. & Strimmer, K. APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20, 289–290 (2004).

    CAS  PubMed  Google Scholar 

  73. Bürkner, P. C. brms: an R package for Bayesian multilevel models using Stan. J. Stat. Softw. 80, 1–28 (2017).

    Google Scholar 

  74. Pinheiro, J. et al. nlme: Linear and nonlinear mixed effects models. R package version 3(1) (2017).

Download references

Acknowledgements

This work was supported by the National Science Foundation through NSF DEB-1556768 (S.J.M. and I.W.P.), NSF DEB-2105932 (S.J.M. and I.W.P.), NSF DEB-2105907 (S.R.) and NSF DEB-2105903 (C.C.D.). T.H.R.-P. is grateful to the University of California, Santa Barbara, for fellowship support in the year this manuscript was completed. We thank the many herbaria, including botanists, staff and volunteers, who collected, curated and digitized the vast volumes of herbarium specimens leveraged for this study. We thank A. Bishop, D. Gamble, C. Hannah-Bick, D. Inouye, L. Kim and H. Payne for comments on earlier drafts of the manuscript.

Author information

Authors and Affiliations

  1. Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, CA, USA

    Tadeo H. Ramirez-Parada, Isaac W. Park & Susan J. Mazer

  2. Department of Wildlife, Fisheries and Conservation Biology, University of Maine, Orono, ME, USA

    Sydne Record

  3. Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, USA

    Charles C. Davis & Aaron M. Ellison

  4. Sound Solutions for Sustainable Science, Boston, MA, USA

    Aaron M. Ellison

Authors
  1. Tadeo H. Ramirez-Parada
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Isaac W. Park
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Sydne Record
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Charles C. Davis
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Aaron M. Ellison
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Susan J. Mazer
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

T.H.R.-P. conceived the initial ideas, which were further developed and refined with S.J.M. and I.W.P. I.W.P. collected the data. T.H.R.-P. designed and conducted the data analyses and created the figures. T.H.R.-P. wrote the first draft and S.J.M., I.W.P., S.R., C.C.D. and A.M.E. contributed significantly to subsequent revisions.

Corresponding author

Correspondence to Tadeo H. Ramirez-Parada.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Ecology & Evolution thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Distributions of and relationship between Sspace and Stime among 201 long-lived species in the continental United States.

Light blue and red shaded regions in (a) respectively correspond to the kernel-density distributions of Sspace and Stime among the 201 species included in this analysis. The solid black line in (b) indicates a 1:1 relationship corresponding to perfect agreement between sensitivity types. The solid curved line indicates the line of best fit obtained from a Generalized Additive Model (GAM) of Stime vs. Sspace, with the shaded area around it denoting the standard error of the predicted mean value. Each point in (b) represents a species whose x, y coordinates are given by the maximum a posteriori (MAP) estimates for Sspace and Stime, respectively. Point shapes and colours in (b) indicate whether sensitivity patterns were consistent with plasticity or adaptation as the sole drivers of flowering time variation along the temperature gradient, with both plasticity and adaptation having statistically significant effects in a co- or counter-gradient adaptation pattern, or not showing statistically significant adaptation nor plasticity. The straight, solid black line in (b) indicates a 1:1 relationship (that is, Sspace = Stime), whereas the curved solid line shows the observed relationship estimated from a generalized additive model (GAM). The shaded region along the curved solid line in (b) corresponds to the standard error of the predicted value of Stime. The percent of species showing each pattern is shown in the legend in parenthesis. The 95% credible interval for the correlation between Sspace and Stime is provided as a text inset in (b). The subset of 201 species was selected based on growth form data from the United States Department of Agriculture Plant Database (USDA Plant Database, https://plants.usda.gov). We downloaded all records of growth habit information available through the search tool and subset the resulting dataset to contain only species represented among the 1,605 species included in the analyses presented in the main text. We then retained only flowering specimens from species classified as ‘Tree’ (n = 5), ‘Shrub’ (n = 164), ‘Subshrub’ (n = 27) or ‘Vine’ (n = 5), which yielded a dataset of 201 species. Using this subset dataset, we ran the model presented in the main text, obtaining estimates of sensitivity to TMEANNormal and TMEANAnomaly and of their difference for each species, as well as an estimate of their correlation accounting for parameter uncertainty. The resulting patterns closely mirrored those of the larger dataset, with a high correlation and agreement in magnitude between Sspace and Stime and similar relative frequencies among species for each sensitivity pattern.

Extended Data Fig. 2 Sampling intensity and long-term climatic conditions across collection sites in the continental United States.

Pixels correspond to 20 × 20-km grid cells, with their colour representing (a) the total number of specimens collected and (b) their mean PC1 and PC2 values. PC1 represents a gradient of increasing precipitation seasonality, decreasing temperature seasonality and increasing long-term mean annual temperature. In turn, PC2 represents a gradient of decreasing long-term mean annual precipitation and increasing temperature seasonality (see ‘Climatic data’).

Extended Data Fig. 3 Estimates with and without detrending DOY and TMEANAnomaly.

Iler et al.64 showed that shared temporal trends between DOY and temperature can generate spurious relationships between these variables that often disappear when the phenological and temperature time series are detrended prior to estimating their relationship. Alternatively, a non-spurious but trended relationship between DOY and temperature might reflect the effects of adaptation to directional changes in temperature, at least in short-lived species. Therefore, relationships between phenology and temperature that persist following detrending are more likely to reflect phenological plasticity. Accordingly, we assessed whether estimates of sensitivity to TMEANAnomaly (Stime) presented in the main text could be confounded by temporal trends in DOY and TMEANAnomaly. To do so, we first ran single-species linear regressions using DOY or TMEANAnomaly as responses and year as a single predictor, storing the resulting residuals as detrended versions of both responses. Then, for each species, we ran two linear models of DOY against TMEANNormal and TMEANAnomaly: one with observed DOY and TMEANAnomaly and another with detrended DOY and TMEANAnomaly. Trended and detrended estimates of sensitivity to TMEANAnomaly were very highly correlated among species, suggesting that TMEAN sensitivity estimates presented in the main text do not reflect the confounding effect of shared temporal trends. Similarly, detrending DOY and TMEANAnomaly did not substantially alter estimates of Sspace and Sspace − Stime. a–c, In each panel, points represent the combinations of trended or detrended estimates of Sspace, Stime, or Sspace − Stime for each species in the data, whereas diagonal black lines correspond to 1:1 to relationships denoting perfect agreement between trended and detrended estimates. Solid blue lines in each panel indicate the observed relationship between trended and detrended estimates, with the shaded region around the trend line (nearly imperceptible due to the large sample size) indicating the standard error of the predicted value.

Extended Data Fig. 4 Variation in Stime across geographic climatic gradients for 1,605 species across the coterminous United States.

Distribution of interaction terms between TMEANAnomaly and long-term climatic conditions within sites of specimen collection, including (a) TMEANNormal, (b) PPT Normal, (c) TMEAN Seasonality, (d) PPT Seasonality, (e) the gradient of increasing temperature and precipitation seasonality described by PC1 and (f) the gradient of decreasing precipitation and increasing temperature seasonality described by PC2. The interaction coefficients for all variables were obtained from single-species models including flowering DOY as a response, the focal long-term climatic variable and TMEANAnomaly as a predictor and an interaction term between them. Long-term climatic variables were standardized (mean = 0, SD = 1) before fitting the models. Accordingly, the interaction terms quantify the change in the slope of TMEANAnomaly vs. DOY (Stime, in days/°C) for an increase of 1 SD in the long-term climatic variable. The values for the 25th, 50th and 75th percentiles of each among-species distribution are indicated in each panel text insets as well as the proportion of species for which the interaction coefficient had a p-value greater than 0.01 (based on a two-sided t-test). Within a species, phenological sensitivity to temperature can vary among portions of the range with different long-term climatic conditions. Therefore, differences between Sspace and Stime presented in the main text may result from variation in the DOY–TMEANAnomaly slope across the geographic climatic gradients and not from the effects of adaptation across the gradient. Despite this, we found no evidence of pervasive variation in Stime along various geographic climatic gradients, suggesting intraspecific variation in phenological sensitivity is unlikely to generate the patterns reported in the main text.

Extended Data Fig. 5 Consequence of including precipitation in models estimating Sspace and Stime.

Relationship between estimates of flowering sensitivity to TMEANNormal (Sspace) (a, c) and TMEANAnomaly (Stime) (b, d) with or without accounting for the effects of cumulative precipitation normal and anomaly (x-axis and y-axis, respectively) during the same 3-month periods used to calculate TMEANNormal and TMEANAnomaly for each species (see Methods in main text). Panels (a) and (b) show the relationships between estimates from temperature-only models with those obtained from models including PPT normal and net PPT anomaly for the focal 3-month period in the year of collection. In turn, panels (c) and (d) show the same relationship but with estimates from a model including PPT normal and PPT anomaly proportional to the long-term average for that period (that is, divided by the PPT normal). Proportional anomalies were included to account for differences in the biological significance that the same amount of precipitation might have in chronically dry compared to chronically wet locations. The method developed by Phillimore et al.9 assumes that the variables causing phenological variation along spatial temperature gradients are correctly identified and included in the model. Although temperature has been found to be a predominant environmental cue inducing flowering in temperate biomes, other variables such as precipitation, or those that emerge from the interaction between temperature and precipitation, such as snow cover or water stress, routinely have been implicated in phenological variation in many North American species. Therefore, it is possible that differences in spatial vs. temporal patterns of temperature-related phenological variation might stem from the confounding effects of phenologically important variables not included in our models. Estimates of phenologically important variables such as the timing of snowmelt or the onset of drought conditions in xeric environments are not available at the temporal and spatial scales spanned by our data. However, most of these variables are highly correlated with precipitation and temperature over space and time and including both in phenoclimatic models might account for the effects of predictors other than temperature and precipitation. Accordingly, we assessed whether estimated Sspace and Stime changed when accounting for the effects of long-term cumulative precipitation (PPT normal) and PPT anomalies in the year of collection, separately assessing the effects of both net PPT anomalies and of anomalies scaled proportionally to long-term means (PPT normal) for the focal 3-month period. Estimates of phenology–temperature relationships in space and time did not change substantially when including precipitation variables, resulting in a very high correlation between estimates from temperature-only models and those from models including precipitation (r = 0.95 or 0.96). Therefore, the estimates presented in the main text are unlikely to be biased by the omission of precipitation during the months leading up to flowering.

Extended Data Fig. 6 Distribution of and relationship between Sspace and Stime among species sampled within narrow latitudinal bands.

These analyses included 157 species with 200 or more specimens collected within a latitudinal band of 1° (~111 km) in the continental United States (analogous to Fig. 2 of the main text). Light blue and red shaded regions in (a) respectively correspond to the kernel-density distributions of Sspace and Stime among the 157 species included in the analysis. The solid black line in (b) indicates a 1:1 relationship corresponding to perfect agreement between the two types of sensitivity. The solid curved line indicates the line of best fit obtained from a Generalized Additive Model (GAM) of Stime vs. Sspace, with the shaded area around it denoting the standard error of the predicted mean value. Each point in (b) represents a species whose x, y coordinates are given by the maximum a posteriori (MAP) estimates for Sspace and Stime, respectively. Point shapes and colours in (b) indicate whether sensitivity patterns were consistent with plasticity or adaptation as the sole drivers of flowering time variation along the temperature gradient, with both plasticity and adaptation having significant effects in a co- or counter-gradient adaptation pattern, or not showing statistically significant adaptation nor plasticity. The straight, solid black line in (b) indicates a 1:1 relationship (that is, Sspace = Stime), whereas the curved solid line shows the observed relationship estimated from a generalized additive model (GAM). The shaded region along the curved solid line in (b) corresponds to the standard error of the predicted value of Stime. The percent of species showing each pattern is shown in the legend in parenthesis. The 95% credible interval for the correlation between Sspace and Stime is provided as a text inset in (b). Both temperature and photoperiod are known to be the predominant environmental cues controlling both vegetative and reproductive phenology among plants in temperature regions. Therefore, across latitudinal ranges such as those spanned by most species in our data (median latitudinal range = ca. 12.2°), it is possible that differences in Stime and Sspace (for example, geographic temperature gradients) might reflect the confounding influence of latitudinal shifts in photoperiod on our estimates of sensitivity to TMEANNormal. To account for this possibility, we identified 157 species in our data that were well sampled (200 or more specimens) within narrow latitudinal bands (≤1°). Using this subset of species and including only specimens from such 1° bands, we ran the model presented in the main text, obtaining estimates of Stime and Sspace and their difference for each species and an estimate of their correlation accounting for parameter uncertainty. The results did not qualitatively differ from those presented in the main text, with a high correlation between Sspace and Stime and similar relative frequencies of each sensitivity pattern among species.

Extended Data Fig. 7 Effects of sample size differences, spatial autocorrelation and phylogeny on estimates of Sspace and Stime.

Comparison of Sspace and Stime estimates obtained by (a) homogenizing sample sizes among species, (b) accounting for spatial autocorrelation among observations and (c) accounting for phylogenetic relationships among species against estimates generated ignoring these factors (as those presented in the main text). In (a, b), we fit the model presented in the main text using a thinned dataset were each species was represented by 300 specimens, comparing its output to that of the model in the main text. In (c, d), we compared the results of models omitting or accounting for phylogenetic relationships. We selected a random subset of 300 species from which to generte a phylogeny, thinning these data to include only 300 specimens for each species (to make the model computationally tractable). Sspace and Stime estimates that did not account for phylogeny were obtained using the model described in the main text. In turn, the model accounting for phylogeny included a prior for the covariance structure of species-specific parameters consisting of the evolutionary distance between each pair of species as estimated from a phylogenetic hypothesis and a model of trait divergence among species. The phylogenetic tree (or hypothesis) was generated using the R package ‘v.PhyloMaker’ version 0.1.071 and generated a phylogeny resovled to the genus level. Using this tree, we then calculated the variance–covariance phylogenetic matrix predicted by a Brownian model of trait evolution using the R package ‘ape’ version 5.6-272. Finally, both models were implemented using the ‘brms’ package version 2.18.073. Finally, in (e, f) we compared estimates obtained from models ignoring or accounting for spatial autocorrelation of the residuals. All Sspace and Stime estimates were obtained using single-species models, but those accounting for spatial autocorrelation included a covariance structure for the residuals determined by the geographic distance between each pair of points. All models were fitted using the’nlme’ package version 3.174 in R. Estimates of Sspace and Stime obtained accounting for or ignoring spatial autocorrelation were nearly indentical across species. Across panels, the x-axes show the estimates obtained when omitting the focal factor (sample size, phylogeny, or spatial autocorrelation), whereas the y-axes show estimates obtained when accounting for it. Solid black lines represent a 1:1 line, representing perfect agreement in magnitude and direction between estimates. Sspace and Stime estimates obtained ignoring sample size differences, phylogeny and spatial autocorrelation where highly correlated to estimates obtained from models accounting for these factors. Accordingly, we consider it unlikely that omitting these factors could have biased our results.

Extended Data Fig. 8 Assessing evidence for nonlinear phenology–temperature relationships.

Comparison of R2 values obtained using 10-fold cross-validation of models of flowering DOY versus TMEANNormal and TMEANAnomaly obtained from (a) linear regressions assuming linear relationships between phenology and temperature or (b) generalized additive models (GAMs) accounting for potential nonlinear relationships. The shaded region in each panel represents the among-species kernel distribution of cross-validated R2 values obtained using each model type (linear regression or GAM). The mean and SD of R2 values each are presented as text insets in each panel. The model that generated the sensitivity estimates presented in the main text assumed linear relationships between flowering dates and TMEANNormal and TMEANAnomaly. To verify whether such an assumption was warranted for our data, we compared the predictive ability of single-species models assuming linear relationships between phenology and temperature (fitted using linear regression) and models accounting for possible nonlinear relationships (fitted using Generalized Additive Models). We reasoned that if omitted nonlinear relationships between flowering time and temperature were pervasive in our data and potentially biased our results, then models accounting for nonlinear relationships would tend to perform better than linear regressions among species in our data. We used 10-fold cross-validation to compare the out-of-sample performance (quantified through R2 values) of linear regressions and GAMs. For each model type (linear regression or GAM), this procedure randomly split the observations for each species into 10 groups, each of which was omitted from a model estimated from the remaining 9 groups. The performance of each of these models was then assessed against the observations omitted in fitting the model, generating 10 out-of-sample R2 values for each model type (linear or GAM) per species. We then compared the distribution of mean cross-validated R2 values obtained from linear models and GAMs to assess whether nonlinear models explained additional variance.

Extended Data Fig. 9 Effects of geographic range on apparent plasticity and adaptation.

Relationships between the latitudinal and longitudinal range of specimens of a species and estimates of apparent plasticity (Stime) and apparent adaptation (Sspace – Stime). Latitudinal ranges in (a, b) and longitudinal ranges in (c, d) were obtained by first removing the extreme 1% of observations among observations for each species. In a–d, blue lines in each panel correspond to best-fit lines obtained using generalized additive models (GAMs), with blue ribbons showing the standard error of the predicted value of the response for each value of the predictors. R2 are provided as text insets in each panel. Although apparent plasticity and adaptation showed marginally greater magnitude among species with narrower latitudinal and longitudinal range, these relationships explained a very small proportion of the variance. Therefore, we conclude that it is unlikely that differences in latitudinal or longitudinal range size could confound the results presented in the main text. GAMs using apparent plasticity or apparent adaptation as a response and including both latitudinal and longitudinal range as predictors also explained a marginal proportion of the variance (R2 = 0.10 and R2 = 0.05, respectively).

Extended Data Fig. 10 Climatic space captured among specimen collection locations across ecoregions.

a–n, Variation in long-term climatic conditions among sites of specimen collection occurring within different Level II ecoregions throughout the contiguous United States. Variation in long-term conditions was calculated using principal components (PCs). PC1 represents a gradient of increasing precipitation seasonality, decreasing temperature seasonality and increasing long-term mean annual temperature. In turn, PC2 represents a gradient of decreasing long-term mean annual precipitation and increasing temperature seasonality (see ‘Climatic data’).

Supplementary information

Supplementary Information

Supplementary Figs. 1–21 and Notes 1–8 with discussion of supplemental analyses.

Reporting Summary

Peer Review File

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ramirez-Parada, T.H., Park, I.W., Record, S. et al. Plasticity and not adaptation is the primary source of temperature-mediated variation in flowering phenology in North America. Nat Ecol Evol 8, 467–476 (2024). https://doi.org/10.1038/s41559-023-02304-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41559-023-02304-5

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing