Advanced Search
Article Contents

Investigation of Uncertainties of Establishment Schemes in Dynamic Global Vegetation Models


doi: 10.1007/s00376-013-3031-1

  • In Dynamic Global Vegetation Models (DGVMs), the establishment of woody vegetation refers to flowering, fertilization, seed production, germination, and the growth of tree seedlings. It determines not only the population densities but also other important ecosystem structural variables. In current DGVMs, establishments of woody plant functional types (PFTs) are assumed to be either the same in the same grid cell, or largely stochastic. We investigated the uncertainties in the competition of establishment among coexisting woody PFTs from three aspects: the dependence of PFT establishments on vegetation states; background establishment; and relative establishment potentials of different PFTs. Sensitivity experiments showed that the dependence of establishment rate on the fractional coverage of a PFT favored the dominant PFT by increasing its share in establishment. While a small background establishment rate had little impact on equilibrium states of the ecosystem, it did change the timescale required for the establishment of alien species in pre-existing forest due to their disadvantage in seed competition during the early stage of invasion. Meanwhile, establishment purely from background (the scheme commonly used in current DGVMs) led to inconsistent behavior in response to the change in PFT specification (e.g., number of PFTs and their specification). Furthermore, the results also indicated that trade-off between individual growth and reproduction/colonization has significant influences on the competition of establishment. Hence, further development of establishment parameterization in DGVMs is essential in reducing the uncertainties in simulations of both ecosystem structures and successions.
    摘要: In Dynamic Global Vegetation Models (DGVMs), the establishment of woody vegetation refers to flowering, fertilization, seed production, germination, and the growth of tree seedlings. It determines not only the population densities but also other important ecosystem structural variables. In current DGVMs, establishments of woody plant functional types (PFTs) are assumed to be either the same in the same grid cell, or largely stochastic. We investigated the uncertainties in the competition of establishment among coexisting woody PFTs from three aspects: the dependence of PFT establishments on vegetation states; background establishment; and relative establishment potentials of different PFTs. Sensitivity experiments showed that the dependence of establishment rate on the fractional coverage of a PFT favored the dominant PFT by increasing its share in establishment. While a small background establishment rate had little impact on equilibrium states of the ecosystem, it did change the timescale required for the establishment of alien species in pre-existing forest due to their disadvantage in seed competition during the early stage of invasion. Meanwhile, establishment purely from background (the scheme commonly used in current DGVMs) led to inconsistent behavior in response to the change in PFT specification (e.g., number of PFTs and their specification). Furthermore, the results also indicated that trade-off between individual growth and reproduction/colonization has significant influences on the competition of establishment. Hence, further development of establishment parameterization in DGVMs is essential in reducing the uncertainties in simulations of both ecosystem structures and successions.
  • 加载中
  • Adams B. , A. White , T. M. Lenton , 2004: An analysis of some diverse approaches to modeling terrestrial net primary productivity. Ecological Modelling, 177, 353- 391.
    Alton P. , L. Mercado , P. North , 2007: A sensitivity analysis of the land-surface scheme JULES conducted for three forest biomes: Biophysical parameters, model processes, and meterorological driving data. Global Biogeochemical Cycles, 20, GB1008, doi: 10.1029/2005GB002653.
    Arai N. , T. Kamitani , 2005: Seed rain and seedling establishment of the dioecious tree Neolitsea sericea (Lauraceae): Effects of tree sex and density on invasion into a conifer plantation in central Japan. Canadian Journal of Botany, 83( 9), 1144- 1150.
    Arora V. K. , G. J. Boer , 2006: Simulating competition and coexistence between plant functional types in a dynamic vegetation model. Earth Interactions, 10( 10), 1- 30.
    Baraloto C. , P. M. Forget , D. E. Goldberg , 2005: Seed mass, seedling size and neotropical tree seedling establishment. Journal of Ecology, 93, 1156- 1166.
    Bonan G. B. , S. Levis , 2006: Evaluating aspects of the community land and atmosphere models (CLM3 and CAM3) using a dynamic global vegetation model. J. Climate, 19, 2290- 2301.
    Bugmann, H. K. M., 1994: On the ecology of mountainous forests in a changing climate: A simulation study. Ph. D. dissertation No. 10638, Swiss Federal Institute of Technology Zurich, Switzerland, 252 pp.
    Bugmann, H. K. M., 2001: A review of forest gap models. Climatic Change, 51, 259- 305.
    Coomes D. A. , P. J. Grubb , 2003: Colonization, tolerance, competition and seed-size variation within functional groups. Trends in Ecology & Evolution, 18, 283- 291.
    Cox P. , 2001: Description of the TRIFFID Dynamic Global Vegetation Model. Hadley Centre Tech. Note 24, Hadley Centre, Bracknell, U.K. , 16 pp.
    Crawford R. M. M. , 2008: Plants at the Margin: Ecological Limits and Climate Change. Cambridge Univ. Press, 478 pp.
    Clark J. S. , M. Lewis , L. Horvath , 2001: Invasion by extremes: Population spread with variation in dispersal and reproduction. The American Naturalist, 157( 5), 537- 554.
    Enright N. J. , B. B. Lamont , 1989: Seed banks, fire season, safe sites and seedling recruitment in five co-occurring Banksia species. Journal of Ecology, 77, 1111- 1122.
    Fargione J. , D. Tilman , 2002: Competition and coexistence in terrestrial plants. U. Sommer and B. Worm, Eds., Competition and Coexistence, Springer-Verlag, Berlin, 165- 207.
    Foley, J. A. , Coauthors, 1996: An integrated biosphere model of land surface processes, terrestrial carbon balance, and vegetation dynamics. Global Biogeochemical Cycles, 10, 603- 628.
    Foley J. A. , S. Levis , I. C. Prentice , D. Pollard , S. L. Thompson , 1998: Coupling dynamic models of climate and vegetation. Global Change Biology, 4, 561- 579.
    Friend A. D. , A. K. Stevens , R. G. Knox , M. G. R. Cannell , 1997: A process-based, terrestrial biosphere model of ecosystem dynamics (Hybrid v3.0). Ecological Modelling, 95, 249- 287.
    Friedlingstein P. , G. Joel , C. B. Field , I. Y. Fung , 1999: Toward an allocation scheme for global terrestrial carbon models. Global Change Biology, 5, 755- 770.
    Friedlingstein, P., Coauthors, 2006: Climate-carbon cycle feedback analysis: Results from the C4MIP model intercomparison. J. Climate, 19, 3337- 3353.
    Gonzalez-Rodriguez V. , R. Villar , R. Casado , E. Suarez-Bonnet, J. L. Quero, R. M. Navarro-Cerrillo , 2011: Spatio-temporal heterogeneity effects on seedling growth and establishment in four Quercus species. Annals of Forest Science, 68, 1217- 1232.
    Grime J. P. , 2001: Plant Strategies, Vegetation Processes, and Ecosystem Properties. John Wiley and Sons Ltd, Chichester, 173 pp.
    Greene D. F. , E. A. Johnson , 1994: Estimating the mean annual seed production of trees. Ecology, 75, 642- 647.
    Haywood A. M. , P. J. Valdes , 2006: Vegetation cover in a warmer world simulated using a dynamic global vegetation model for the Mid-Pliocene. Palaeogeography Palaeoclimatology Palaeoecology, 237, 412- 427.
    Hughes J. K. , P. J. Valdes , R. A. Betts , 2004: Dynamical properties of the TRIFFID dynamic global vegetation model. Hadley Centre Tech. Note 56, Hadley Centre, Exeter, U.K. , 23 pp.
    Iponga D. M. , S. J. Milton , D. M. Richardson , 2009: Reproductive potential and seedling establishment of the invasive alien tree Schinus molle (Anacardiaceae) in South Africa. Austral Ecology, 34( 6), 678- 687.
    Khurana E. , J. S. Singh , 2001: Ecology of seed and seedling growth for conservation and restoration of tropical dry forest: a review. Environmental Conservation, 28( 1), 39- 52.
    Knorr W. , M. Heimann , 2001: Uncertainties in global terrestrial biosphere modelling. 1. A comprehensive sensitivity analysis with a new photosynthesis and energy balance scheme. Global 370 Biogeochemical Cycles, 15, 201- 225.
    Krinner, G., Coauthors, 2005: A dynamic global vegetation model for studies of the coupled atmosphere-biosphere system. Global Biogeochemical Cycles, 19, GB1015, doi: 10.1029/2003GB002199.
    Kucharik, C. J. , Coauthors, 2000: Testing the performance of a Dynamic Global Ecosystem Model: Water balance, carbon balance, and vegetation structure. Global Biogeochemical Cycles, 14, 795- 825.
    Levis S. , G. B. Bonan , M. Vertenstein , K.W. Oleson , 2004: The Community Land Model’s Dynamic Global Vegetation Model (CLM-DGVM): Technical description and user’s guide. NCAR Tech. Note, NCAR/TN-459+IA, National Center for Atmospheric Research, Boulder, NCAR, Colorado, 50 pp.
    Lewis S. L. , 2006: Tropical forests and the changing earth system. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 361, 195- 210, doi: 10.1098/rstb.2005.1711.
    Mao J. F. , B.Wang, Y. J. Dai , F. I.Woodward , P. J. Hanson , M. R. Lomas , 2007: Improvements of a Dynamic Global Vegetation Model and simulations of carbon and water at an upland-oak forest. Adv. Atmos. Sci., 24, 311- 322, doi: 10.1007/s00376-007-0311-7.
    McCloy K. R. , W. Lucht , 2004: Comparative evaluation of seasonal patterns in long time series of satellite image data and simulations of a global vegetation model. IEEE Trans. Geosci. Remote Sens., 42, 140- 153.
    Meir P. , P. Cox , J. Grace , 2006: The influence of terrestrial ecosystems on climate. Trends in Ecology & Evolution, 21, 254- 260.
    Milberg P. , B. B. Lamont , 1997: Seed/cotyledon size and nutrient content play a major role in early performance of species on nutrient-poor soil. New Phytologist, 137, 665- 672.
    Moles A. T. , M. Westoby , 2004: Seedling survival and seed size: A synthesis of the literature. Journal of Ecology, 92, 372- 383.
    Moorcroft P. R. , 2003: Recent advances in ecosystem-atmosphere interactions: An ecological perspective. Philosophical Transactions of the Royal Society of London Series B-Biological Sciences, 270, 1215- 1227, doi: 10.1098/rspb.2002.2251.
    Moorcroft P. R. , G. C. Hurtt , S. W. Pacala , 2001: A method for scaling vegetation dynamics: The ecosystem demography model (ED). Ecological Monographs, 71( 4), 557- 586.
    Oleson, K. W. , Coauthors, 2004: Technical description of the Community Land Model (CLM). NCAR Tech. Note, NCAR/TN-461+STR, National Center for Atmospheric Research, Boulder, Colorado, 174 pp.
    Paz H. , S. J. Mazer , M. Martinez-Ramos , 1999: Seed mass, seedling emergence, and environmental factors in seven rain forest Psychotoria (Rubiaceae). Ecology, 80( 5), 1594- 1606.
    Prentice I. C. , T. M. Sykes , W. Cramer , 1993: A simulation model for the transient effects of climate change on forest landscapes. Ecological Modelling, 65, 51- 70.
    Qian T. T. , A. G. Dai , K. E. Trenberth , K. W. Oleson , 2006: Simulation of global land surface conditions from 1948 to 2004. Part I: Forcing data and evaluations. Journal of Hydrometeorology, 7, 953- 975.
    Richardson D. M. , 1998: Forestry tree as invasive aliens. Conservation Biology, 12( 1), 18- 26, doi: 10.1046/j.1523-1739.1998.96392.x.
    Rey P. J. , J. M. Alcántara, F. Valera, A. M. Sánchez-Lafuente, J. L. Garrido, J. M. Ramirez , A. J. Manzaneda , 2004: Seedling establishment in Olea europaea: Seed size and microhabitat affect growth and survival. Ecoscience, 11( 1), 310- 320.
    Sato H. , A. Itoh , T. Kohyama , 2007: SEIB-DGVM: A new Dynamic Global Vegetation Model using a spatially explicit individual-based approach. Ecological Modelling, 200, 279- 307.
    Sitch, S., Coauthors, 2003: Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biology, 9, 161- 185.
    Sitch, S., Coauthors, 2008: Evaluation of the terrestrial carbon cycle, future plant geography and climate-carbon cycle feedbacks using five Dynamic Global Vegetation Models (DGVMs). Global Change Biology, 14, 2015- 2039.
    Smith B. , I. C. Prentice , M. T. Sykes , 2001: Representation of vegetation dynamics in the modelling of terrestrial ecosystems: Comparing two contrasting approaches within European climate space. Global Ecology Biogeography, 10, 621- 637.
    Stoy, P. C. , Coauthors, 2008: Role of vegetation in determining carbon sequestration along ecological succession in the southeastern United States. Global Change Biology, 14, 1- 19, doi: 10.1111/j.1365-4252486.2008.01587.x.
    Urbieta I. R. , I. M. Perez-Ramos, M. A. Zavala, T. Maranon , R. K. Kobe , 2008: Soil water content and emergence time control seedling establishment in three co-occurring Mediterranean oak species. Canadian Journal of Forest Research, 38, 2382- 2393.
    Vazquez-Yanes C. , A. Orozco-Segovia , 1992: Effects of litter from a tropical rainforest on tree seed germination and establishment under controlled conditions. Tree Physiology, 11, 391- 400.
    Vieira D. L. M. , A. Scariot , 2006: Principles of natural regeneration of tropical dry forests for restoration. Restoration Ecology, 14( 1), 11- 20.
    Weiner J. , 1988: The influence of competition on plant reproduction. Plant Reproduction Ecology: Patterns and Strategies, J. Lovett Doust and L. Lovett Doust, Eds., Oxford Univ. Press, Oxford, 228- 245.
    Wramneby A. , B. Smith , S. Zaehle , M. T. Sykes , 2008: Parameter uncertainties in the modeling of vegetation dynamics-Effects on tree community structure and ecosystem functioning in European forest biomes. Ecological Modelling, 216, 277- 290.
    Yang D. W. , W. W. Shao , P. J. F. Yeh , H. B. Yang , S. Kanae , T. Oki , 2009: Impact of vegetation coverage on regional water balance in the nonhumid regions of China. Water Resources Research, 45, W00A14, doi: 10.1029/2008WR006948.
    Zeng X. D. , 2010: Evaluating the dependence of vegetation on climate in an improved Dynamic Global Vegetation Model (CLM3.0-DGVM). Adv. Atmos. Sci., 27, 977- 991, doi: 10.1007/s00376-009-9186-0.
    Zeng X. D. , X. B. Zeng , M. Barlage , 2008: Growing temperate shrubs over arid and semiarid regions in the Community Land Model-Dynamic Global Vegetation Model. Global Biogeochemical Cycles, 22, GB3003, doi: 10.1029/2007GB003014.
  • [1] Xiang SONG, Xiaodong ZENG, Jiawen ZHU, Pu SHAO, 2016: Development of an Establishment Scheme for a DGVM, ADVANCES IN ATMOSPHERIC SCIENCES, 33, 829-840.  doi: 10.1007/s00376-016-5284-y
    [2] ZENG Xiaodong, LI Fang, SONG Xiang, 2014: Development of the IAP Dynamic Global Vegetation Model, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 505-514.  doi: 10.1007/s00376-013-3155-3
    [3] HU Shujuan, CHOU Jifan, 2004: Uncertainty of the Numerical Solution of a Nonlinear System's Long-term Behavior and Global Convergence of the Numerical Pattern, ADVANCES IN ATMOSPHERIC SCIENCES, 21, 767-774.  doi: 10.1007/BF02916373
    [4] ZENG Xiaodong, 2010: Evaluating the Dependence of Vegetation on Climate in an Improved Dynamic Global Vegetation Model, ADVANCES IN ATMOSPHERIC SCIENCES, 27, 977-991.  doi: 10.1007/s00376-009-9186-0
    [5] Jiawen ZHU, Xiaodong ZENG, Minghua ZHANG, Yongjiu DAI, Duoying JI, Fang LI, Qian ZHANG, He ZHANG, Xiang SONG, 2018: Evaluation of the New Dynamic Global Vegetation Model in CAS-ESM, ADVANCES IN ATMOSPHERIC SCIENCES, 35, 659-670.  doi: 10.1007/s00376-017-7154-7
    [6] LI Fang, ZENG Xiaodong, SONG Xiang, TIAN Dongxiao, SHAO Pu, ZHANG Dongling, 2011: Impact of Spin-up Forcing on Vegetation States Simulated by a Dynamic Global Vegetation Model Coupled with a Land Surface Model, ADVANCES IN ATMOSPHERIC SCIENCES, 28, 775-788.  doi: 10.1007/s00376-010-0009-0
    [7] Chenxi WANG, Zhihua ZENG, Ming YING, 2020: Uncertainty in Tropical Cyclone Intensity Predictions due to Uncertainty in Initial Conditions, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 278-290.  doi: 10.1007/s00376-019-9126-6
    [8] JIANG Dabang, 2008: Projected Potential Vegetation Change in China under the SRES A2 and B2 Scenarios, ADVANCES IN ATMOSPHERIC SCIENCES, 25, 126-138.  doi: 10.1007/s00376-008-0126-1
    [9] Xiaoxin WANG, Dabang JIANG, Xianmei LANG, 2018: Climate Change of 4°C Global Warming above Pre-industrial Levels, ADVANCES IN ATMOSPHERIC SCIENCES, 35, 757-770.  doi: 10.1007/s00376-018-7160-4
    [10] Jianguo LIU, Binghao JIA, Zhenghui XIE, Chunxiang SHI, 2016: Ensemble Simulation of Land Evapotranspiration in China Based on a Multi-Forcing and Multi-Model Approach, ADVANCES IN ATMOSPHERIC SCIENCES, 33, 673-684.  doi: 10.1007/s00376-016-5213-0
    [11] Jianguo LIU, Zong-Liang YANG, Binghao JIA, Longhuan WANG, Ping WANG, Zhenghui XIE, Chunxiang SHI, 2023: Elucidating Dominant Factors Affecting Land Surface Hydrological Simulations of the Community Land Model over China, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 235-250.  doi: 10.1007/s00376-022-2091-5
    [12] Deliang CHEN, Christine ACHBERGER, Jouni R¨AIS¨ANEN, Cecilia HELLSTR¨OM, 2006: Using Statistical Downscaling to Quantify the GCM-Related Uncertainty in Regional Climate Change Scenarios: A Case Study of Swedish Precipitation, ADVANCES IN ATMOSPHERIC SCIENCES, 23, 54-60.  doi: 10.1007/s00376-006-0006-5
    [13] Yuejian ZHU, 2005: Ensemble Forecast: A New Approach to Uncertainty and Predictability, ADVANCES IN ATMOSPHERIC SCIENCES, 22, 781-788.  doi: 10.1007/BF02918678
    [14] SONG Xiang, ZENG Xiaodong, ZHU Jiawen, 2013: Evaluating the Tree Population Density and Its Impacts in CLM-DGVM, ADVANCES IN ATMOSPHERIC SCIENCES, 30, 116-124.  doi: 10.1007/s00376-012-1271-0
    [15] MAO Jiafu, DAN Li, WANG Bin, DAI Yongjiu, 2010: Simulation and Evaluation of Terrestrial Ecosystem NPP with M-SDGVM over Continental China, ADVANCES IN ATMOSPHERIC SCIENCES, 27, 427-442.  doi: 10.1007/s00376--009-9006--6
    [16] ZHOU Mengzi, WANG Huijun, 2015: Potential Impact of Future Climate Change on Crop Yield in Northeastern China, ADVANCES IN ATMOSPHERIC SCIENCES, 32, 889-897.  doi: 10.1007/s00376-014-4161-9
    [17] TIAN Di, GUO Yan*, DONG Wenjie, 2015: Future Changes and Uncertainties in Temperature and Precipitation over China Based on CMIP5 Models, ADVANCES IN ATMOSPHERIC SCIENCES, 32, 487-496.  doi: 10.1007/s00376-014-4102-7
    [18] Lin WANG, Gang HUANG, Wen ZHOU, Wen CHEN, 2016: Historical Change and Future Scenarios of Sea Level Rise in Macau and Adjacent Waters, ADVANCES IN ATMOSPHERIC SCIENCES, 33, 462-475.  doi: 10.1007/s00376-015-5047-1
    [19] Guoxiong WU, Bian HE, Anmin DUAN, Yimin LIU, Wei YU, 2017: Formation and Variation of the Atmospheric Heat Source over the Tibetan Plateau and Its Climate Effects, ADVANCES IN ATMOSPHERIC SCIENCES, 34, 1169-1184.  doi: 10.1007/s00376-017-7014-5
    [20] GUO Yanjun, DING Yihui, 2011: Impacts of Reference Time Series on the Homogenization of Radiosonde Temperature, ADVANCES IN ATMOSPHERIC SCIENCES, 28, 1011-1022.  doi: 10.1007/s00376-010-9211-3

Get Citation+

Export:  

Share Article

Manuscript History

Manuscript received: 04 February 2013
Manuscript revised: 24 April 2013
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Investigation of Uncertainties of Establishment Schemes in Dynamic Global Vegetation Models

  • 1. International Center for Climate and Environment Sciences, Institute of Atmospheric Physics,Chinese Academy of Sciences, Beijing 100029

Abstract: In Dynamic Global Vegetation Models (DGVMs), the establishment of woody vegetation refers to flowering, fertilization, seed production, germination, and the growth of tree seedlings. It determines not only the population densities but also other important ecosystem structural variables. In current DGVMs, establishments of woody plant functional types (PFTs) are assumed to be either the same in the same grid cell, or largely stochastic. We investigated the uncertainties in the competition of establishment among coexisting woody PFTs from three aspects: the dependence of PFT establishments on vegetation states; background establishment; and relative establishment potentials of different PFTs. Sensitivity experiments showed that the dependence of establishment rate on the fractional coverage of a PFT favored the dominant PFT by increasing its share in establishment. While a small background establishment rate had little impact on equilibrium states of the ecosystem, it did change the timescale required for the establishment of alien species in pre-existing forest due to their disadvantage in seed competition during the early stage of invasion. Meanwhile, establishment purely from background (the scheme commonly used in current DGVMs) led to inconsistent behavior in response to the change in PFT specification (e.g., number of PFTs and their specification). Furthermore, the results also indicated that trade-off between individual growth and reproduction/colonization has significant influences on the competition of establishment. Hence, further development of establishment parameterization in DGVMs is essential in reducing the uncertainties in simulations of both ecosystem structures and successions.

摘要: In Dynamic Global Vegetation Models (DGVMs), the establishment of woody vegetation refers to flowering, fertilization, seed production, germination, and the growth of tree seedlings. It determines not only the population densities but also other important ecosystem structural variables. In current DGVMs, establishments of woody plant functional types (PFTs) are assumed to be either the same in the same grid cell, or largely stochastic. We investigated the uncertainties in the competition of establishment among coexisting woody PFTs from three aspects: the dependence of PFT establishments on vegetation states; background establishment; and relative establishment potentials of different PFTs. Sensitivity experiments showed that the dependence of establishment rate on the fractional coverage of a PFT favored the dominant PFT by increasing its share in establishment. While a small background establishment rate had little impact on equilibrium states of the ecosystem, it did change the timescale required for the establishment of alien species in pre-existing forest due to their disadvantage in seed competition during the early stage of invasion. Meanwhile, establishment purely from background (the scheme commonly used in current DGVMs) led to inconsistent behavior in response to the change in PFT specification (e.g., number of PFTs and their specification). Furthermore, the results also indicated that trade-off between individual growth and reproduction/colonization has significant influences on the competition of establishment. Hence, further development of establishment parameterization in DGVMs is essential in reducing the uncertainties in simulations of both ecosystem structures and successions.

1 Introduction
  • Vegetation is one of the most important elements of the Earth’s biosphere. It works as a CO2 sink, a modulator of hydrologic flow, and has a very close relationship with climate change through water circulation and energy cycles (Moorcroft, 2003; Friedlingstein et al., 2006; Lewis, 2006; Meir et al., 2006; Stoy et al., 2008). In recent years, various vegetation simulation models have been developed to study terrestrial ecosystem processes and the relationship between climate change and vegetation; of these, Dynamic Global Vegetation Models (DGVMs) are important members (Foley et al., 1998; Smith et al., 2001; Sitch et al., 2003; Levis et al., 2004; Krinner et al., 2005; Haywood and Valdes, 2006; Mao et al., 2007; Sato et al., 2007; Sitch et al., 2008). So far, most DGVMs can reproduce the global distribution of natural vegetation under the current climate conditions, and capture the relationship between natural vegetation distribution and climate conditions (Kucharik et al., 2000;McCloy and Lucht, 2004). However, the most important mission of DGVMs is to predict the transient responses of ecosystem distribution and structure to climate change (Friend et al., 1997; Moorcroft et al., 2001; Zeng, 2010). Unfortunately, current DGVMs are still in their infancy and hence do not work well as predictive tools. One of the biggest problems is the uncertainties in models, which may be introduced by uncertainties in physics sub-models (including imprecise physics, data-driven empirical models) (Knorr and Heimann, 2001; Alton et al., 2007) and excessive simplification of parameterizations. In this context, we discuss the parameterization of establishment in the present paper.

    In nature, the well-known process that includes seed production, propagation, storage, and the establishment of saplings, is a uniquely effective way to reproduce for most trees and shrubs, and it plays a key role in community structure and transitions [Grime, 2001]. Generally, a failure to grow or an inability to reproduce/establish are the two major limitations to vegetation distribution [Crawford, 2008]. Competition-colonization trade-off is an important mechanism of vegetation coexistence [Fargione and Tilman, 2002], especially in resource-limiting regions, where vegetation with a greater capacity for seed production may have an advantage in occupying open space, at least during the early stage of colonization. The successfulness of establishment of a species/plant functional type (PFT) is affected by various factors, such as climate and environmental factors (light, soil moisture, temperature, growing-degree days, aboveground litter etc.) (Enright and Lamont, 1989; Vazquez-Yanes and Orozco-Segovia, 1992; Vieira and Scariot, 2006; Urbieta et al., 2008; Gonzalez-Rodriguez et al., 2011), the physiological traits of plants (seed yields, size of seeds etc.) (Paz et al., 1999; Rey et al., 2004; Baraloto et al., 2005), the current state of the vegetation (e.g., biomass or percentage of coverage), competition among surrounding seedlings, and tolerance of shading from already-established vegetation [Coomes and Grubb, 2003].

    DGVMs consider the major biogeophysical, biogeochemical, and ecological processes on individual and population levels. Much attention has been focused on individual growth, such as formulations of photosynthesis [Adams et al., 2004], allocation of net primary production (Foley et al., 1996; Friend et al., 1997; Friedlingstein et al., 1999), competition for light and water (Cox, 2001; Sitch et al., 2003; Hughes et al., 2004; Krinner et al., 2005; Arora and Boer, 2006), and so on. However, processes at the population level, especially establishment and mortality, have not been well formulated, despite these processes having great effects on vegetation distribution [Bugmann, 2001]. In a given model, establishment may refer to a sequence of biological processes including flowering, fertilization, seed production, germination, and the establishment of new individuals. When the establishment rate equals the death rate, the tree population density will be close to steady state. Furthermore, competition for establishment from different PFTs also plays an important role in ecosystem structure and transition.

    Existing DGVMs possess various establishment schemes, in which the real establishment process is simplified based on one or several physical or ecological mechanisms. Two steps are usually included: (1) the species or PFTs that can establish in the given year (or reference time) are determined by the current environmental conditions, such as precipitation, light availability at the forest floor, temperature, and growing-degree days; and (2) the increment of population density, i.e., individuals for each species or PFT to be established, is calculated. Some models, such as LPJ (Lund-Potsdam-Jena Dynamic Global Vegetation Model) (Sitch et al., 2003, 2008), CLM-DGVM (Community Land Model-Dynamic Global Vegetation Model) [Levis et al., 2004], SDGVM (Sheffield Dynamic Global Vegetation Model) [Sitch et al., 2008] suppose that the population density increment of the established PFTs is proportional to the available area, and each established woody PFT has the same population increment. SEIB-DGVM (Spatially Explicit Individual-based Dynamic Global Vegetation Model) has four different establishment schemes for woody PFTs. One of them is similar to that of LPJ, i.e., every woody PFT that meets the establishment conditions shares the available mesh box equally, while another scenario is that only one woody PFT can monopolize the available space. The other two scenarios are a little more complicated. In the first of these, it is assumed that, to begin with, every established woody PFT has the same establishment rate; then, after a given year, the available mesh box allocated among woody PFTs is proportional to their biomass. In the second of these last two scenarios, long-distance seed dispersal is considered based on the third approach. TRIFFID (Top-down Representation of Interactive Foliage and Flora Including Dynamics) [Cox, 2001] is special in that there is no separate establishment sub-model and does not have the definition of population density for PFTs. Instead, in each year, a minimum “seed” fraction for all PFTs is added in the carbon balance equation to ensure their establishment for each time step. Similar to TRIFFID, CTEM (Canadian Terrestrial Ecosystem Model) [Arora and Boer, 2006] simulates the change in community structure with modified Lotka-Volterra equations. However, it assumes that, as long as the climate permits, each PFT has the potential of establishment, which depends on colonization rates, i.e., a PFT without a “parent” can also be established successfully. This point is guaranteed by the traits of the modified Lotka-Volterra equations. In the ED model (Ecosystem Demography Model) [Moorcroft et al., 2001], the recruitment of new seedlings is governed by a Neumann boundary condition of a partial differential equation, and the number of seedlings is determined by the given PFT’s net primary production, the probability of seed germination and survival, as well as the size of seeds.

    Unlike DGVMs, forest gap models usually use a random number with uniform [Bugmann, 1994] or Poisson (Prentice et al., 1993; Smith et al., 2001; Wramneby et al., 2008) distribution to calculate the number of new individuals, and the traits of distributions are related to species characteristics (e.g., the maximum sapling establishment rate and some physiology traits) and/or the maximum sapling establishment rate, patch size, environmental conditions (e.g., light conditions at the forest floor, soil moisture etc.).

    The differences in the above establishment parameterization schemes may have impacts not only on the global distribution of each PFT, but also may influence the PFT fractional coverage and variability in each grid cell through population dynamics. Considering that the biogeographical constraints delineating PFT establishment can be somehow inferred from field observations and are not very different among DGVMs, the major uncertainties of establishment schemes exist in three aspects: (1) the total amount of individuals to be established from all PFTs in a grid cell; (2) the partitioning of this total establishment to different PFTs; and (3) the proportion of established seedlings that can eventually mature. With the aim to address the uncertainty concerning the second of these three aspects, a revised CLM-DGVM is used as the test bed. The paper is organized as follows. Section 2 introduces the revised CLM3.0-DGVM, focusing in particular on its establishment sub-model, and then a more general form of establishment scheme is brought forth. This is followed by sensitivity tests and analysis in section 3, and further discussion and conclusions are presented finally in section 4.

2 Description of the revised CLM3.0-DGVM and a more general scheme for establishment
  • CLM3.0-DGVM refers to the Dynamic Global Vegetation Model [Levis et al., 2004] of the Community Land Model 3.0 [Oleson et al., 2004]. It considers photosynthesis, respiration, phenology, allocation, competition, survival and establishment, mortality, litter decomposition, soil respiration, as well as fire disturbance. Different processes are calculated on different time scales varying from twenty minutes to one year. In this paper, a revised version of CLM3.0-DGVM (Zeng et al., 2008; Zeng, 2010) is used. The revised model involves a sub-model of temperate and boreal shrubs, a “two-leaf” scheme for photosynthesis, and a new definition of fractional coverage of PFTs. The default and revised CLM-DGVM have been used to simulate the global distribution of forest, shrub land, grassland, and desert (Bonan and Levis, 2006; Zeng et al., 2008), as well as roughly reproduce corresponding relations between vegetation distribution and climate [Zeng, 2010].

    PFT fractional coverage (F) is one of the most important vegetation state variables in evaluating a DGVM’s performance. It has remarkable impacts on surface processes, such as regional water balance [Yang et al., 2009], soil water, temperature and so forth. As a mean field model, CLM-DGVM ignores the difference between individuals within the same PFT, and calculates F of woody PFTs as [Zeng et al., 2008]

    Fi=σiPi (1)

    where σi (m2) and Pi (individuals m-2) are the averaged crown area and population density of the ith woody PFT, respectively. In the model, P is a prognostic variable, and its change (ΔP) in a given time step (i.e., 1 year) is determined by

    ΔPPestPlightPmortPfire (2)

    Where ΔPest is the establishment rate, and ΔPlight, ΔPmort, ΔPfire are the population decreases due to light competition, mortality and fire, respectively. Meanwhile, σ is a diagnostic variable calculated from the tissue biomasses according to allometric relationships, and it is worth noting that the tissue biomasses are averaged among the grown-up and newly-established individuals.

    Thus, establishment has two direct impacts on the dynamic behaviors of the system. First, at steady states, establishment counterbalances the other decreasing terms in Eq. (2). Second, during the initial spin-up period and disturbance period, when the size differences between saplings and established individuals for a given tree PFT are small, the establishment can increase the individual number as well as the PFT total biomasses rapidly, and thereby increase the fractional coverage.

    In the following, we focus on the establishment scheme of CLM3.0-DGVM and introduce the new scheme.

  • The revised model uses the same establishment scheme as the default CLM-DGVM, except that the newly added shrub PFTs are treated as woody PFTs and follow the same formula of establishment (the default model cannot simulate shrub PFTs).

    CLM3.0-DGVM has two steps in its establishment scheme. First, a given woody PFT can establish in a grid cell if the required climate conditions are satisfied, e.g., Tc,min < Tc < Tc,max, GDD 5> GDD min, GDD23 equals to zero, and the annual precipitation is greater than 100 mm. Here, Tc refers to the 20-yr running mean of the minimum monthly temperature, GDD5 and GDD23 are the 20-yr running means of the annual growing-degree days above 5°C and 23°C, respectively. The parameters Tc,min , Tc,max, GDDmin are PFT-dependent constants.

    Second, the establishment rate of a given woody PFT, ΔPi, is calculated once a year as [Eq. (53) of [Levis et al., 2004]]

    Where ΔPmax is the maximum establishment rate, which equals 0.24 individuals m-2yr-1; Fwoody is the total fractional coverage of woody PFTs in the grid cell; the term (1- Fwoody) is the space that is not occupied by the existing woody PFTs; the term refers to the shading effects [Sitch et al., 2003]; the term denotes the total establishment rate for all woody PFTs in the given grid cell; nest,woody refers to the number of established woody PFTs in the grid cell in the current year; and the term 1/nest,woody indicates that all the established woody PFTs have the same establishment capability. Without loss of generality, 1/nest,woody can be replaced by

    where gi denotes the weight of the ith woody PFT in establishment competition, and Eq. (3) corresponds to the case of

    gi≡1 (5)

    Therefore, a more general establishment scheme can be formulated by

  • Equation (5) (i.e., the scheme in CLM-DGVM) assumes that all woody PFTs have the same establishment rate. It neglects the current states of the PFTs, such as fractional coverage, net primary production (NPP), the number of seeds and so forth. To solve such an unrealistic aspect in the establishment of different PFTs, the following three aspects should be incorporated: (1) Generally speaking, the photosynthate allocated for reproduction is related to the amount of canopy area for photosynthesis (Weiner, 1988; Greene and Johnson, 1994;Moles andWestoby, 2004). So, a PFT with higher fractional coverage would reach suitable germination sites more often than other PFTs, and hence has more chance of establishment in a given area. However, individuals of the same PFT usually tend to grow in clusters; seeds produced in the central area of a PFT distribution are less likely to be succeeded in establishment due to shading effects and resource competition, while the seeds produced in exterior areas find it easier to reach suitable sites. Hence, the relationship between PFT fractional coverage and establishment rate might be nonlinear. (2) Besides the establishment from seeds produced in the current year by local existing PFTs, there may also be settlement from seeds produced in the previous year and/or invasion from surrounding grid cells may also lead to establishment in the receiving grid cell, which can be treated as background establishment. This is important in the model because it ensures the re-establishment of temporally extinguished species. (3) Different PFTs may have different establishment capacity. For example, some species, especially those with relatively lower productivity in the region might evolve a competition strategy in that they allocate a higher amount of their NPP to reproduction, so that they can produce relatively larger amount of seeds (i.e., higher establishment rate) and hence quickly expand and outnumber the PFT with higher productivity but a lower reproductive rate. Therefore, a more general scheme can be formulated as

    gi= gi0[ε0 +(1-ε0)Fiα], (7)

    where gi0 is the PFT-dependent constant of relative establishment potential, i.e., the establishment capability of the ith PFT when Fi=100%; Fi is the fractional coverage of the ith woody PFT in the current year; α<1 is an exponential factor representing the cluster effects; and ε0=1 is a constant denoting background establishment.

    It is worth pointing out that Eq. (5) can be considered as the extreme case of Eq. (7) with α≡0 and gi0≡1. Actually, in the early stage (i.e., spin-up) when Fs of all woody PFTs are small, the background establishment plays the dominant role in the establishment rate, and the scheme in Eq. (7) acts more or less as Eq. (5). However, when Fs of one or a few woody PFTs gradually become larger, these PFTs also become dominant in overall establishment, making establishment for other PFTs more difficult.

3 Sensitivity tests with the revised CLM3.0-DGVM
  • Because there are not enough observational data to determine the parameters (α, ε0 and gi0) in Eq. (7), sensitivity experiments were performed to investigate the uncertainties of the establishment scheme by evaluating the impacts of these parameters on the structure of simulated ecosystems. The revised CLM-DGVM was used as the test bed. First, global offline simulations, running 800-1500 years and forced circularly with 50 years of reanalysis surface atmospheric fields (1950-99) from [Qian et al., 2006] were performed using the default establishment scheme in CLM-DGVM [i.e., Eq. (3)]. A grid cell located in the Indo-China Peninsula [centered at (14.3°N, 105.0°E)] was selected, which satisfied two conditions, i.e., (a) tree PFTs were dominant, and (b) there were at least two tree PFTs with F>10%, allowing the establishment competition between different tree PFTs to be depicted. In this grid cell, the mean annual precipitation and temperature was about 1600 mm and 27°C, respectively. Such humid tropical climate conditions satisfied the growth conditions of BET-Tr (tropical broadleaf evergreen tree), BDT-Tr (tropical broadleaf deciduous tree) and C4 grass, with their Fs being about 25%, 52% and 23%, respectively. There also existed BDM-Sh (temperate broadleaf deciduous shrub) with F less than 0.1%, and hence was omitted from the results.

    Figure 1.  Variation of the establishment rate (Pest, individuals m-2yr-1), population density (P, individuals m-2), and fractional coverage (F, %) in the 800-yr single-point simulations. In (a)-(c), α=0, while in (d)-(l) α=1/3, 1/2, and 1, respectively. The center of the grid cell is (14.3°N, 105.0°E).

    Figure 2.  Comparison between α=0 and α=0.5 in the grid cell (14.3°N, 105.0°E). ΔFtree (%) denotes the difference in the total fractional coverage summed over tree PFTs with BDM-Sh removed; ΔFdiff (%) denotes the change in the difference between the fractional coverage of BDT-Tr and BET-Tr with BDM-Sh removed.

    Figure 3.  Impacts of the potential establishment capability (gi0) of PFTs on F and P with α=0.5, where gi0=1 and PCT reprod=10% for all woody PFTs, but gi0=2, PCTreprod=15% for BET-Tr in (a) and (b), and gi0=4, PCT reprod=20% for BET-Tr in (c) and (d). The center of the grid cell is the same as in Fig. 1.

    Sensitivity tests with different values of α,ε0 and gi0 were then performed on the selected grid cell. For simplification, the base values of the parameters were chosen as α=0.5, ε0=0.01, and gi0=1, and only one parameter was changed in each set of experiments. For comparison, in each set there was one experiment with α=0 (orε0=1) and gi0=1, which corresponded to the default establishment scheme in CLM-DGVM.

  • Figure 1 shows the results of 800-yr single-point simulations. When α=0, all woody PFTs that met the climate conditions in the grid cell were able to establish and had the same establishment rate (Fig. 1a), even if some PFTs had negative NPPs. Consequently, the differences among population densities of existing woody PFTs were relatively small (Fig. 1b). However, when α==1/3,1/2, and 1, the increment of sapling individuals was proportional to its Fα, so the dominant woody PFT with larger F had the advantage in establishment (Figs. 1d, g and j), and hence had the highest population density in the grid cell (Figs. 1e, h and k), which further enhanced the difference in F between the dominant woody PFT and other PFTs, with the difference becoming larger as α increased.

  • Usually, there are more than two woody PFTs existing within a grid cell. Some of them have very low F s and hence their impacts on ecosystem dynamics should be very small. However, this is not the case when α=0. For example, when all woody PFTs with F s <1% were removed from the same grid cell as used for the experiment described in Section 3.1, and only two dominant tree PFTs (BET-Tr and BDT-Tr) and C4 grass were allowed to establish and survive, significant differences in ecosystem composition were found when α=0 (Fig. 2), in which the total tree coverage (Ftree) increased by about 8%, and the difference between the F s of BET-Tr and BDT-Tr (Fdiff) also increased slightly, by 1%. Such unrealistic behavior is caused by the assumption that woody PFTs in a grid cell share the same establishment rate in Eq. (5). However, when α=0.5, the removal of PFTs resulted in slight differences in both Ftree and Fdiff during the spin-up period, but such differences gradually attenuated to almost 0 after 200 years of simulation.

    Figure 4.  Comparison of the time scale of PFT invasions betweenα=0 andα=0.5 with different ε0, where year 0 in the figure corresponds to the year when BDT-Tr is introduced and BET-Tr as well as C4 grass reach equilibrium. In (a) and (b), α=0 and in (c) and (f) α=0.5, with ε0=0.1 and 0.01, respectively. F(%) is the fractional coverage of PFTs, and P (individuals m-2) is the population density of PFTs.

  • In the simulations with Eq. (7) described above, all the tree PFTs were assumed to apply the same establishment strategy, e.g., invest the same percentage of annual NPP in establishment (PCTreprod, 10%). In order to depict different establishment strategies for different PFTs, two sensitivity simulations using Eq. (7) were performed on the same grid cell as used for the experiment described in section 3.1, with gi0=2 and 4 for BET-Tr (the second dominant tree PFT when α=0) respectively, while gi0=1 for all other woody PFTs. Meanwhile, PCTreprod for BET-Tr was increased to 15% and 20%, respectively. Compared with the results shown in Figs. 1h and i, both P and F of BET-Tr increased significantly (Fig. 3). It took much longer for BDT-Tr to exceed BET-Tr (Fig. 3a), i.e., 475 years cf. 150 years (Fig. 1i), and BET-Tr eventually became dominant (Fig. 3c), by limiting the P of BDT-Tr to a relatively lower level (Fig. 3d).

  • Plenty of observational work has shown that the number of seeds and the survival probabilities of seeds and saplings play critical roles in species invasion (Richardson, 1998; Clark et al., 2001; Arai and Kamitani, 2005; Iponga et al., 2009). In order to investigate the effects from different establishment schemes on the time scale of PFT invasions, two sensitivity simulations were performed on grid cell (14.3°N, 105.0°E). First, only the second dominant tree PFT (BET-Tr) and C4 grass were allowed to establish with α=0. Then, after the model reached the point of approaching steady state, the first original dominant tree PFT (BDT-Tr) was added, with α=0 and 0.5 respectively, to simulate the effects of invasion from alien species. The results showed that α=0.5 slowed the invasion rate of BDT-Tr (Figs. 4c and e vs. Fig. 4a) by slowing its growth of P (Fig. 4d and f vs. Fig. 4b). The time taken for BDT-Tr to become dominant was almost 370 years when α=0.5 and ε0=0.1 (Fig. 4c), compared with about 250 years when α=0, and the time scale was lengthened as ε0 became smaller (370 years vs. 450 years). BET-Tr was the absolutely dominant PFT with about 80% F at the beginning. Hence, when α=0.5, the seed production, spread and establishment rate for the invader (BDT-Tr) remained at a low level for a long period. However, whenα=0, the invasive species had the same establishment rate with local dominant PFTs, and hence its P increased quickly.

4 Discussion and conclusions
  • The uncertainties in competition of establishment among woody PFTs in DGVMs have been investigated. A revised version of CLM-DGVM was used as the test platform. The main conclusions of the study are: (1) The dominant PFT may take the advantage in competition of establishment with the dependence of establishment rate on vegetation states, but coexistence of different PFTs may be possible if the cluster effects factor [i.e., α in Eq. (7)] is small. (2) While a small background establishment rate (ε0) mainly influences the timescale required for the establishment of alien species, a largeε0 (as implied in the scheme commonly used in current DGVMs) may lead to inconsistent model behaviors, e.g., changes in the total tree fractional coverages in response to the removal or adding of locally rare species. (3) Furthermore, trade-off between individual growth and reproduction/colonization will have a significant influence on the competition of establishment.

    For simplicity, background establishment rate is assumed to be spatially homogenous and independent of species in most DGVMs. Our results imply that background establishment rate, ε0, has a significant influence only on PFTs with low fractional coverage, e.g., during the recovery or establishment of invading species. Hence, major uncertainties caused byε0 may come from the marginal areas. For example, within the ecotone (transition zone between different ecosystems), competition among species is strong, and the ecosystem is sensitive to climate and environmental changes. Population densities are usually low compared with central areas of the corresponding PFT’s distribution, and hence seeds propagated from neighboring areas are very important in maintaining species population and may have large variabilities. Another special case is island effects. Being far away from large continents, the background establishment rates for locally non-existing PFTs should be nearly 0, and hence the response of the island ecosystem to climate change would be quite different. These situations should be considered in the parameterization of establishment.

    The strategy of individual growth and reproduction trade-off by different PFTs is known to be very important in vegetation competition, but is usually ignored in most DGVMs. This is reflected by the establishment potentials gi0 in this paper. Besides, there are also differences in seed traits, such as seed weight and size, propagation properties, shade tolerance, mortality rate, and so on. For example, some PFTs produces a large number of small seeds that disperse widely and germinate faster but with higher mortality, while others have smaller seeds, and such seeds usually have greater viability and lower mortality when subjected to external disturbances or environmental changes (Milberg and Lamont, 1997; Khurana and Singh, 2001; Rey et al., 2004). Such details may have impacts on both the equilibrium of ecosystem states as well as an ecosystem’s response to changes in climate and the environment. Therefore, how to introduce the above details into establishment schemes, and how influential they are, require further investigation.

Reference

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return