Elsevier

Ecological Modelling

Volume 394, 24 February 2019, Pages 34-43
Ecological Modelling

Modelling leaf coloration dates over temperate China by considering effects of leafy season climate

https://doi.org/10.1016/j.ecolmodel.2018.12.020Get rights and content

Highlights

  • The leaf senescence models are improved by incorporating effects of leafy season climate.

  • Leafy season thermal and moisture conditions are major climate drivers in addition to autumn temperature and photoperiod.

  • Local adaptation of leaf senescence cannot be detected by current process-based models.

  • Climatic effects might be stronger than genetic effects.

Abstract

Current process-based models of autumn phenophases are generally based on autumn temperature and/or photoperiod cues. The dependence of autumn phenology on environmental conditions occurring throughout the leafy season has been overlooked. In this study, we incorporated the effect of leafy season temperature and precipitation in process-based models with the aim to improve the modelling of autumn phenology. We tested the ability of three existing and three new autumn phenology models in predicting the occurrence of leaf coloration of four deciduous tree species during 1981–2012 across temperate China. The results show that the models taking the effects of both the leafy season temperature and low precipitation into account performed best over both calibration and validation data. Compared with existing autumn phenological models, the best models predict that the potential delay of autumn phenology in a warmer world is modulated by the impact of leafy season climate on leaf senescence in a complex way: increased leafy season temperature tends to hasten the occurrence of leaf coloration while low leafy season precipitation tends to delay it. Additionally, we tested the hypothesis of a local adaptation of tree phenology through the evaluation of site-specific model parameterizations. Local fittings of the critical temperature sums in all models did not improve the model ability to simulate the occurrence of leaf coloration, which suggests either that local adaptation of leaf senescence process is virtually non-existent across the populations considered or that the genetic variation of leaf coloration traits among populations cannot be detected by our models. Our findings highlighted the importance of leafy season climate in autumn phenology modelling and its possible offset effect in the response of autumn senescence to future warming.

Introduction

Plant phenology regulates ecosystem processes related to the carbon (Delpierre et al., 2009b), water (Hogg et al., 2000) and nutrient (Cooke and Weih, 2005; Escudero and Del Arco, 1987; Miki and Doi, 2016) cycles, and feeds back to the climate system (Peñuelas and Filella, 2009; Richardson et al., 2013; Schwartz, 1992). An accurate representation of plant phenology is essential for terrestrial ecosystem models, of which some are used as land surface schemes for Earth system models (Richardson et al., 2012). Much effort has been done for the modelling of spring phenological phases (Chuine, 2000; Chuine et al., 2013; Hänninen, 1995; Kramer, 1994; Xu and Chen, 2013), but the modelling of autumn phenophases like leaf coloration and leaf fall (Delpierre et al., 2009a; Keenan and Richardson, 2015) has received less attention. Thus, there are still relatively few autumn phenology models, and most display a limited ability to accurately predict the timing of autumn phenophases, hence their responses to climate change (Gallinat et al., 2015; Gill et al., 2015).

For temperate and boreal tree species, several statistical and process-based models have been developed to simulate the leaf coloration dates (Delpierre et al., 2009a; Dufrêne et al., 2005; Jeong and Medvigy, 2014; Jolly et al., 2005; Keenan and Richardson, 2015; Olsson and Jönsson, 2015; Richardson et al., 2006; White et al., 1997; Xie et al., 2018). Although statistical models that rely on correlations between the timing of autumn phenological events and climatic cues (Chuine et al., 2013; Hanninen and Kramer, 2007) are simple and sometimes may perform better than process-based models (Olsson and Jönsson, 2014), they are considered not to be applicable out of their calibration range, which hinders their application for predicting autumn phenology under future climate change. Therefore, process-based models are urgently needed.

To date, few process-based models have been developed for simulating leaf coloration (Delpierre et al., 2009a; Dufrêne et al., 2005; Keenan and Richardson, 2015). These models share a common framework in which the leaf coloration process was defined as three components (Delpierre et al., 2009a) : 1) the start date of leaf senescence processes, that is the timing when mature leaves begin entering into senescence state; 2) the daily leaf senescence development rate, which represents the influence of climatic factors on the progress of leaf coloration; 3) a required “critical value” (typically, a cold-temperature sum threshold) for the occurrence of leaf coloration. The start date of leaf senescence process is usually set on a fixed date such as summer solstice (Yu et al., 2016) or is determined by a photoperiod threshold (Delpierre et al., 2009a; Jeong and Medvigy, 2014; Keenan and Richardson, 2015). That means leaf senescence was proposed to occur after a fixed date or when photoperiod dropped below a species-specific threshold. The description of leaf senescence development rate is a crucial part in leaf coloration modelling. Mathematically, it is a function of climatic factors influencing leaf senescence. In some process-based models (Dufrêne et al., 2005; Yu et al., 2016), temperature is considered as the only factor to affect leaf coloration progress, in which the leaf coloration date is proposed to occur when the accumulated cold-degree-day sum (CDD) fulfils a given requirement. Yet, some studies evidenced that, along with temperature, photoperiod is another important factor influencing leaf coloration (Delpierre et al., 2009a; Estiarte and Peñuelas, 2015; Fracheboud et al., 2009). A process-based model combining the effect of both temperature and photoperiod on leaf coloration has been developed by Delpierre et al. (2009a), and applied widely over the temperate zone (Archetti et al., 2013; Tao et al., 2018; Vitasse et al., 2011; Yang et al., 2012). This model assumes that, after the summer solstice, shortening photoperiod starts triggering the initiation of leaf senescence, which is subsequently accelerated by low temperature, possibly in interaction with photoperiod. Generally, the third component – the required critical cold-temperature sum for forcing leaf coloration is considered to be constant for a given species, being considered as a cross-population average phenological trait (Delpierre et al., 2009a; Dufrêne et al., 2005). Based on this framework, the spring-influenced autumn phenology model (SIAM) (Keenan and Richardson, 2015) modified the required critical cold-temperature sum to assume it varying with spring budburst, in line with the relations between spring and autumn phenology (earlier/later spring budburst could lead to earlier/later leaf coloration) reported by some studies (Fu et al., 2014; Keenan and Richardson, 2015).

In boreal and temperate trees, the annual cycle of organ growth and senescence is a continuous process and forms an integrated system where one phase in the cycle affects the subsequent phases (Delpierre et al., 2016; Hanninen and Kramer, 2007; Hänninen and Tanino, 2011; Sarvas, 1972, 1974). Accordingly, in line with the SIAM model, one may assume that both the summer and autumn climatic conditions regulate the occurrence of autumn leaf coloration. Some evidence reported a negative correlation between the leaf coloration and temperature conditions prevailing several months earlier (e.g. May–June) (Archetti et al., 2013; Estrella and Menzel, 2006; Gordo and Sanz, 2010), which suggests that previous climate experienced by trees may impact the leaf coloration date. In a previous work, we further demonstrated that beyond autumn climatic conditions, temperature experienced by the trees during leafy season also influences the date of leaf coloration (Liu et al., 2018): higher leafy season temperature is associated with earlier leaf coloration date beside the promoting influence of autumn low temperature, which has been confirmed from an experimental study (Fu et al., 2018). According to such results, we may expect the critical cold-temperature sum at which leaf coloration occurs to be smaller following a warm growing season. In addition, environmental stresses such as drought (Hwang et al., 2014; Xie et al., 2015) or nutrient deficiency can also hasten the timing of leaf coloration (Gan and Amasino, 1997; Lim et al., 2007). Recent studies found that meteorological drought (a precipitation shortfall) would delay the occurrence of leaf coloration (Xie et al., 2015, 2018), which is opposite to what is expected under ecophysiological drought (Bréda et al., 2006; Lim et al., 2007). Studies show that drought events have increased worldwide (Dai, 2011; Trenberth et al., 2014), and northeastern China has experienced severe and prolonged dry period (Li et al., 2015). However, there is no process-based model considering drought effect on leaf coloration to date.

In this study, we used long-term ground observations of leaf phenology for four deciduous forest tree species collected from 1981 to 2012 in temperate China and aimed to evaluate and simulate the effects of leafy season climate on leaf coloration dates through a process-oriented modelling approach. We proposed three new models incorporating the effect of leafy season climate (namely leafy season temperature, low precipitation index and their interaction) on leaf coloration and compared their abilities with three previously published models. We further tested the hypothesis of a local adaptation of the autumn phenology of tree populations through the evaluation of species-specific vs. site-specific model parameterizations.

Section snippets

Phenological observations

We focused on four major deciduous tree species with wide distribution in China: Hankow willow (Salix matsudana Koidz.), Siberian elm (Ulmus pumila L.), Black locust (Robinia pseudoacacia L.), and Simon poplar (Populus simonii Carr.). The phenological observations for the four tree species were obtained from the countrywide phenological monitoring network of China Meterological Administration (Chen, 2013). We selected the first leaf unfolding (LU), and full leaf coloration (LC) observations in

Species-specific modelling comparisons

In all study species, ME values above zero in both calibration and validation datasets show that all models outperformed the Null model (the latter being defined as the mean observed LC date for each species). Among the six LC models considered, the model integrating an effect of Tls and LPIls on the required critical cold-temperature sum (i.e. TPDM, Table 1) showed the best performances with minimum AICc over calibration data for all species except Black locust. When tested on the validation

Model comparison and climatic factors in leaf coloration models

Various hypotheses about the climatic controls on leaf senescence have been proposed (Estrella and Menzel, 2006), but they have not been translated to mechanistic models with good predictive power. Currently, most autumn phenological models hypothesize leaf senescence to be the controlled by late-summer and /or autumn temperature and day length variation (Delpierre et al., 2009a; Fu et al., 2018; White et al., 1997). In our previous study (Liu et al., 2018), we completed this view reporting

Conflicts of interests

The authors declare no conflicts of interests.

Acknowledgements

The authors thank the Meteorological Information Center of the China Meteorological Administration for providing field phenological data and the phenological group of Peking University for digitizing the phenological records. We thank Isabelle Chuine for valuable comments and suggestions. We acknowledge support from the National Natural Science Foundation of China under grant no. 41471033, 41771049 and 31770516. Guohua LIU also appreciates the CSC (China Scholarship Council) for a doctoral

References (77)

  • A.D. Richardson et al.

    Climate change, phenology, and phenological control of vegetation feedbacks to the climate system

    Agric. For. Meteorol.

    (2013)
  • Z. Tao et al.

    Modeling spatiotemporal variations in leaf coloring date of three tree species across China

    Agric. For. Meteorol.

    (2018)
  • Y. Vitasse et al.

    Assessing the effects of climate change on the phenology of European temperate trees

    Agric. For. Meteorol.

    (2011)
  • Y. Xie et al.

    Predicting autumn phenology: how deciduous tree species respond to weather stressors

    Agr. Forest Meteorol.

    (2018)
  • M. Archetti et al.

    Predicting climate change impacts on the amount and duration of autumn colors in a New England forest

    PLoS One

    (2013)
  • N. Bréda et al.

    Temperate forest trees and stands under severe drought: a review of ecophysiological responses, adaptation processes and long-term consequences

    Ann. For. Sci.

    (2006)
  • X. Chen

    East Asia, Phenology: An Integrative Environmental Science

    (2013)
  • China Meterological Administration

    Phenological Observation (in Chinese), Observation Criterion of Agicultural Meteorology (in Chinese)

    (1993)
  • I. Chuine et al.

    Fitting models predicting dates of flowering of temperate-zone trees using simulated annealing

    Plant Cell Environ.

    (1998)
  • I. Chuine et al.

    Selecting models to predict the timing of flowering of temperate trees: implications for tree phenology modelling

    Plant Cell Environ.

    (1999)
  • I. Chuine et al.

    Plant Development Models, Phenology: An Integrative Environmental Science

    (2013)
  • J.E. Cooke et al.

    Nitrogen storage and seasonal nitrogen cycling in Populus: bridging molecular physiology and ecophysiology

    New Phytol.

    (2005)
  • A. Dai

    Drought under global warming: a review

    Wiley Interdiscip. Rev. Clim. Change

    (2011)
  • N. Delpierre et al.

    Exceptional carbon uptake in European forests during the warm spring of 2007: a data–model analysis

    Global Change Biol.

    (2009)
  • N. Delpierre et al.

    Temperate and boreal forest tree phenology: from organ-scale processes to terrestrial ecosystem models

    Ann. For. Sci.

    (2016)
  • A. Escudero et al.

    Ecological significance of the phenology of leaf abscission

    Oikos

    (1987)
  • M. Estiarte et al.

    Alteration of the phenology of leaf senescence and fall in winter deciduous species by climate change: effects on nutrient proficiency

    Global Change Biol.

    (2015)
  • N. Estrella et al.

    Responses of leaf colouring in four deciduous tree species to climate and weather in Germany

    Clim. Chang. Res. LettRes.

    (2006)
  • Y. Fracheboud et al.

    The control of autumn senescence in European aspen

    Plant Physiol.

    (2009)
  • Y.S. Fu et al.

    Variation in leaf flushing date influences autumnal senescence and next year’s flushing date in two temperate tree species

    Proc. Natl. Acad. Sci.

    (2014)
  • Y.H. Fu et al.

    Larger temperature response of autumn leaf senescence than spring leaf-out phenology

    Global Change Biol.

    (2018)
  • S. Gan et al.

    Making sense of senescence (molecular genetic regulation and manipulation of leaf senescence)

    Plant Physiol.

    (1997)
  • A.L. Gill et al.

    Changes in autumn senescence in northern hemisphere deciduous trees: a meta-analysis of autumn phenology studies

    Ann. Bot.

    (2015)
  • O. Gordo et al.

    Impact of climate change on plant phenology in Mediterranean ecosystems

    Global Change Biol.

    (2010)
  • C.A. Gunderson et al.

    Forest phenology and a warmer climate–growing season extension in relation to climatic provenance

    Global Change Biol.

    (2012)
  • G. Guyot

    Physics of the Environment and Climate

    (1998)
  • H. Hänninen

    Effects of climatic change on trees from cool and temperate regions: an ecophysiological approach to modelling of bud burst phenology

    Can. J. Bot.

    (1995)
  • H. Hänninen

    Boreal and Temperate Trees in a Changing Climate

    (2016)
  • Cited by (22)

    • Phenological shifts induced by climate change amplify drought for broad-leaved trees at low elevations in Switzerland

      2021, Agricultural and Forest Meteorology
      Citation Excerpt :

      Consequently, phenological shifts amplified drought intensity after 1985 at low elevations for beech, rowan, and sycamore, but weakened it at high elevations for beech, rowan, sycamore, and larch. Assuming that these trends will continue in the future (Delpierre et al. 2009, Keenan and Richardson 2015, Liu et al. 2019; but see Zani et al. 2020), populations of these species are likely to be impaired by drought, especially at lower elevation. Signs of drought impacts are already clearly visible and widespread at lowlands in Switzerland for several deciduous species, particularly for beech in response to recent summer droughts (Arend et al. 2016, Schuldt et al. 2020).

    • Leafy season length is reduced by a prolonged soil water deficit but not by repeated defoliation in beech trees (Fagus sylvatica L.): comparison of response among regional populations grown in a common garden

      2021, Agricultural and Forest Meteorology
      Citation Excerpt :

      Another more drastic rainfall exclusion experiment reducing rainfall by 50% in autumn showed no impact on leaf development the following spring for Quercus ilex whereas a 58% reduction in rainfall during spring led to severe aborting of buds (Misson et al., 2011). How soil water deficits interact with warm temperature in leaf senescence is complex; the two parameters have yet to be disentangled (Estiarte and Peñuelas, 2015; Liu et al., 2019). Escudero and Del Arco (1987) showed that soil water deficit induced earlier leaf fall, but Pallardy and Loewenstein (2004) observed that this response was species dependent.

    • Changes in sessile oak (Quercus petraea) productivity under climate change by improved leaf phenology in the 3-PG model

      2020, Ecological Modelling
      Citation Excerpt :

      Yet, several studies also report the lack of a clear trend for changes in leaf senescence dates (Schaber and Badeck, 2005; Menzel et al., 2008). Leaf senescence processes are yet not very well understood and have reported to be influenced by many different factors, such as drought conditions (Vitasse et al., 2010; Bréda et al., 2006; Pataki and Oren, 2003) or drought in combination with elevated CO2 (Asshoff et al., 2006; Warren et al., 2011), temperatures and precipitation previous to autumn (Liu et al., 2019), coupling with spring phenology (Keenan and Richardson, 2015; Fu et al., 2014), nutrient availability (Sigurdsson, 2001) or photoperiod (Delpierre et al., 2009). 3-PG accounts for water limitations with a reduction in productivity as a consequence to a reduction in available soil water.

    View all citing articles on Scopus
    View full text