Experimental warming differentially affects vegetative and reproductive phenology of tundra plants

Rapid climate warming is altering Arctic and alpine tundra ecosystem structure and function, including shifts in plant phenology. While the advancement of green up and flowering are well-documented, it remains unclear whether all phenophases, particularly those later in the season, will shift in unison or respond divergently to warming. Here, we present the largest synthesis to our knowledge of experimental warming effects on tundra plant phenology from the International Tundra Experiment. We examine the effect of warming on a suite of season-wide plant phenophases. Results challenge the expectation that all phenophases will advance in unison to warming. Instead, we find that experimental warming caused: (1) larger phenological shifts in reproductive versus vegetative phenophases and (2) advanced reproductive phenophases and green up but delayed leaf senescence which translated to a lengthening of the growing season by approximately 3%. Patterns were consistent across sites, plant species and over time. The advancement of reproductive seasons and lengthening of growing seasons may have significant consequences for trophic interactions and ecosystem function across the tundra.


nature research | reporting summary
April 2020 Field-specific reporting Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf Ecological, evolutionary & environmental sciences study design All studies must disclose on these points even when the disclosure is negative.

Research sample
Sampling strategy

Data collection
Timing and spatial scale

Reproducibility
This study is a quantitative synthesis of tundra plant phenology responses to long-term climate warming manipulations, based on 46 experiments conducted at 18 sites in alpine and Arctic locations worldwide, with observations of over 118 plant species spanning from 1992-2019 as part of the International Tundra Experiment. We incorporate multiple plant phenophases across the entire growing season (green-up, flowering, end-of-flowering, fruiting, seed-dispersal, and leaf-senescence) to assess how warming differentially impacts reproductive versus vegetative phenology, as well as early versus late season phenology. Experimental warming is carried out at experimental locations within each site using passive open-top warming chambers (OTCs) made of fiberglass in either cone or hexagon shape approximately 1.5-2 m in diameter. We use interval-censored modeling and Bayesian hierarchical modeling to control for sources of variation within the data including: differences in sampling intervals, site location, subsite location (e.g. experimental location nested within site), plant species identity, and time of sampling (year nested within site). We also examine interactions between experimental warming (i.e. treatment) and the following covariates: 1) Years of warming (continuous, replicate level), 2) latitude (continuous, site level), 3) water availability (categorical:dry/moist/wet) based on gravimetric water content (GWC), site:subsite level), 4) OTC deployment period (categorical (year-round/summer only), site level), 5) site mean temperature (continuous, site level) and 6) site-year temperature anomaly (continuous, site:year level). Number of observations, sites, subsites, years and species for each plant phenophase and number of replicates included in each hierarchical model can be found in Table 2.
Research sample includes recorded dates of phenological events (e.g. first flowering date, first leaf color change) for each plant species being monitored at a given site at either the individual plant or plot level. We chose this sample type in order to understand the influence of warming on the timing of plant phenology. Samples are intended to represent a population of each monitored plant species at each experimental location within a site. Phenology observations were taken using a common protocol outlined in the ITEX manual (Molau, U. & MØlgaard, 1996), yet sites included slightly different phenology definitions across sub-sites (experimental plots) and species (e.g. flower open vs. bud break), and we included whichever phenophase definition was most commonly measured at a given sub-site for each species across all years. We then grouped these measurements across all sites and categorized them as one of the six standardized phenophases above. All site and species specific phenology definitions can be found in Appendix S2.
Following the ITEX protocol, observers recorded the phenological status of plants one to three times per week over the snow-free season. This frequency allowed for estimation of start and end dates of different phenophases without putting undue burden on researchers (i.e., taking phenology observations every single day would be too time consuming and logistically not feasible at some sites). We used interval censored modeling to control for differences in sampling intervals across sites. Minimum sample size in the field was 20 experimentally warmed individuals (plants) and 20 control plants (across all species) for each monitoring subsite in a given sampling year as indicated in the ITEX manual. In hierarchical models we required at least two measurements of each species in both OTCs and control plots in a given site x subsite x year combination. Rationale for this cutoff follows the advice of Gelman and Hill (2007) that 'even two observations per group is enough to fit a multilevel model' (p 276).
Site PIs, Postdocs, graduate students and/or field technicians recorded the phenological status of plants through visual, in person observations one to three times per week over the snow-free season. Measurements were recorded in the field using standardized field data sheets for each species available in the appendix of the ITEX manual (pg IX-XXI).
Depending on site, data collection started as early as 1992, through as recent as 2019. Years of data collection varied across sites due to differences in funding, and the specific years included at each site can be found in Table 1 and Appendix S3d. Following the ITEX protocol, observers recorded the phenological status of plants one to three times per week over the snow-free season (approximately May 1-September 1). This frequency allowed for estimation of start and end dates of different phenophases without putting undue burden on researchers (i.e., taking phenology observations every single day would be too time consuming and logistically not feasible at some sites). We used interval censored modeling to control for differences in sampling intervals across sites. The spatial scale of this analysis is shown in Fig 2 and the spatial scale of sample collection was the size of open-top warming chambers (OTCs) 1.5-2 m in diameter and uniformly sized control plots (1.5-2m x 1.5-2m).
All exclusion criteria were established before analyses began. We excluded any species x sub-site x year combination where more than 20% of the total observations were NAs, for green-up and leaf-senescence only, because missing data can bias effect sizes and because these phenophases at the beginning and end of the growing season are particularly prone to missing data, as the phenological event may have already occurred before the first visit date or may have occurred after the last visit date. Prior to regression, we also discarded any spp x subsite x year combinations that did not have at least two observations in both OTC and control treatments and removed outliers where the difference in OTC vs. control was greater than 4 standard deviations from the mean for that phenophase.
This is a data synthesis, with field monitoring data collected over time, so we can not reproduce the experiments per se. However all all data and code for analyses from this synthesis will be made publicly available so that others may use our analytical approach with their own phenological data.