Phenological mismatches between above- and below-ground plant responses to climate warming: a global synthesis

23 Climate warming is changing above-ground phenology of plants around the world 1, 2 . However, warming effects on below-ground phenology of plants are 25 unclear despite that roots play a vital role in carbon cycling 3 . By conducting a 26 global meta-analysis, we show a phenological mismatch between above- and 27 below-ground plant responses to climate warming. Herbaceous plants advanced 28 both the start and end of the growing season based on their above-ground 29 responses, resulting into a shorter growing season. Below-ground phenophases 30 did not exhibit any obvious changes in herbaceous plants. In contrast, climate 31 warming did not affect the length of above-ground growing season but extended 32 the below-ground growing season of woody plants. These results highlight that 33 climate warming can differentially affect above- and below-ground plant 34 phenology with mismatches arising in herbaceous plants via less responsive 35 below-ground phenology whereas mismatches in woody plants via more 36 responsive below-ground phenology. Mismatches in above- and below-ground 37 plant phenology imply that terrestrial carbon cycling models exclusively based 38 on above-ground responses are less accurate, which highlight the urgent need to 39 incorporate below-ground plant phenology into future Earth system models.

responses to warming, we suspect that our current predictions of ecosystem processes 68 (e.g., soil carbon dynamics) under climate change scenario merely using above-69 ground phenology could be lesser reliable. Here we aim to unravel how climate warming alters both above-and below-ground   82 Climate warming significantly advanced the start and the end of the growing season 83 by 2.88 (95% CI: -1.85~-3.91) and 5.1 days (95% CI: -2.84~-7.11) based on above-     (Table S1). 99 For herbaceous plants, the start of the growing season was not significantly affected 100 by any examined factors based on above-ground responses (Fig. 3A). However, the 101 response of the end of the above-ground growing season was strongly affected by that For woody plants, we found that the start of the growing season based on above-111 ground response was clearly affected by experimental duration (Fig. 3B), with a 112 greater advancement observed over the experimental duration (Fig. 4D). However, the 113 below-ground phenophases of woody plants were not influenced by any factors 114 examined in our study (Fig. 3). global quantitative synthesis. First, we found that below-ground phenology responds 121 differently to warming among different functional groups, that is, below-ground 122 phenology of herbaceous plants was less sensitive to warming than woody plants. 123 Second, we demonstrate that even within the same trophic group (i.e., plants),

189
Data compilation 190 We used meta-analysis to assess the effect of experimental warming on above-and  Table S3).

210
Ancillary site information including latitude, longitude, annual mean air temperature, 211 and mean annual precipitation were also compiled. Annual mean air temperature and 212 annual precipitation were taken directly from original papers or from the cited papers.

213
If these data were not presented, we extracted them from the "worldclim" database 214 (www.worldclim.com). 216 Similar to previous studies 29, 30 , the start and the end of the above-ground growing 217 season were defined as the Julian days at which 10% and 90% of annual growth in 218 above-ground dry matter, stem height, stem diameter or leaf biomass were 219 accumulated, while those of the below-ground growing season were defined as the 220 Julian days when 10% and 90% of annual growth in root dry matter or root length 221 were accumulated. Here, we used the production between the first sampling time and   Egger's regression and Fail-Safe Analysis were used to test the publication bias (Table   269 S2). All the statistical analysis was conducted by R 3.6.1.  The global distribution of warming experiments selected in this meta analysis. The blue and red circles indicate which experiment provided seasonality data of herbaceous and woody plants, respectively. Note:

Phenological parameters extraction
The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.

Figure 2
Response of above-and below-ground phenological events to warming in herbaceous (A, C, E) and woody (B, D, F) plants. SOS, EOS and length represent start, end and length of growing season, respectively. Error bars indicate 95% bootstrapped con dence intervals (CI). The vertical dotted lines are drawn at effect size equals 0. The warming effect was considered signi cant if the 95% CI of the effect size did not overlap with zero. The observation numbers were shown in brackets. *, P < 0.05; **, P < 0.01; ***, P<0.001.  Contributors identi ed from model selection for the warming-induced shifts in above-and below-ground phenophases in herbaceous (A-C) and woody (D) plants. The size of the circles was proportional to the weights of the observation. The relationships between the warming-induced shifts in the start of aboveground growing season and warming-induced shifts in the end of above-ground growing season in herbaceous plants (A). The relationships between experimental duration and warming-induced shifts in start of below-ground growing season (B). The relationships between MAP and warming-induced shifts in start of below-ground growing season (C). The relationships between experimental duration and warming323 induced shifts in start of above-ground growing season (D). SOS, start of the growing season; EOS, end of the growing season.