Elsevier

Ecological Modelling

Volume 319, 10 January 2016, Pages 233-254
Ecological Modelling

Analysing the spatial and temporal dynamics of species interactions in mixed-species forests and the effects of stand density using the 3-PG model

https://doi.org/10.1016/j.ecolmodel.2015.07.010Get rights and content

Highlights

  • The 3-PG forest model was modified for mixed-species forests and deciduous species.

  • 3-PGmix was tested in subtropical monocultures and mixed-species forests in China.

  • 3-PGmix predicted biomass, partitioning and light absorption.

  • 3-PGmix simulated effects of fertility, climate, age and density on species interactions.

Abstract

The growth dynamics and ecosystem services from mixed-species stands are often difficult to predict because the way a given combination of species interacts changes as resource availability or climatic conditions change from site to site or as stands develop over time. Empirical data for many of these situations is often nonexistent. The forest growth model 3-PG was adapted for mixed-species forests and for thinned stands by modifying the light-absorption routine and allowing for within-canopy vertical gradients in climate in the water balance routine. It was also adapted for deciduous species and to predict diameter distributions. The resulting model, 3-PGmix, was used to examine the growth dynamics of subtropical mixed-species forests containing Castanopsis sclerophylla, Cunninghamia lanceolata and Liquidambar formosana, with a wide range of stand densities in Shitai County, Anhui Province, China. After parameterizing and calibrating 3-PGmix using data from monocultures, its predictions of leaf and stem biomass, basal area and light absorption by each species within the mixture were highly correlated with measured values. 3-PGmix also predicted spatial and temporal changes in complementarity, expressed as the relative differences in growth of a given species in mixture compared with its monoculture. Complementarity was predicted to change as the stands developed, across gradients in fertility and rainfall, and also at different stand densities. Such information could be used to suggest potential species compositions, species proportions and thinning regimes for new mixed-species plantations or the responses of existing mixtures to changes in species proportions, stand density and climate. 3-PG is widely used as a routine management tool for monospecific stands and this study shows that it could potentially be used for mixed-species stands as well.

Introduction

Mixed-species forests can be more productive than monocultures when complementary traits and species interactions enable them to use a higher proportion of the sites resources or to use them more efficiently. However, mixtures are certainly not always more productive than monocultures. A given combination of species can be more productive than monocultures on some sites or ages but less productive elsewhere (Forrester, 2014a). This is because the interactions between species changes spatially and temporally with changes in resource availability or climatic conditions (Forrester, 2014a). Species interactions can also be modified by differences in stand density (Garber and Maguire, 2004, Amoroso and Turnblom, 2006, Condés et al., 2013) and probably also by management activities or disturbances that influence density. To make the most of these interactions and to determine how to manage forests under changing management and climatic conditions, tools that can predict these spatial and temporal dynamics and disturbance effects are required.

There are countless combinations of species, silvicultural treatments, soil and climatic conditions in any given region. In many cases, foresters are interested in combinations of these variables that may not currently exist, especially when the goal is to establish new mixed-species plantations. Traditional empirical growth and yield models often unrealistically assume stable site conditions (Battaglia and Sands, 1998). In some tree-level models inter-specific interactions are estimated from empirical relationships based on competition indices calculated from the size of, or shading from, neighbouring trees (Peng, 2000, Pretzsch et al., 2015). These empirical relationships often require large data sets to develop and are restricted to the stand structures or site and climatic conditions represented in the data set (Battaglia and Sands, 1998, Peng, 2000, Landsberg, 2003). This is because they do not consider the main factors that influence forest growth including resource availability, climatic conditions or different silvicultural regimes (Korzukhin et al., 1996, Monserud, 2003). On the other hand, models that are based on general and fundamental ecophysiological processes can potentially provide robust extrapolations to untested conditions, silvicultural regimes (Weiskittel et al., 2010) and species combinations and proportions. While some important physiological processes work at high temporal (seconds, hours) or spatial (leaves, trees) resolutions, models that operate at these resolutions often require detailed and expensive physiological data to parameterise and validate (Battaglia and Sands, 1998). Calculations at high resolutions can also lead to errors that are propagated when upscaling, and are not necessary when the desired outputs are at lower temporal (months or years) or spatial (stands) resolutions, such as those often required by forest managers.

In an attempt to bridge the gap between conventional empirical models and process-based models, hybrid models have been developed to capture some of the advantages of both (Landsberg and Waring, 1997, Battaglia and Sands, 1998, Sands and Landsberg, 2002, Landsberg, 2003, Härkönen et al., 2010, Landsberg and Sands, 2010, Weiskittel et al., 2010). These include the generality and ability of process-based models to describe the forests’ interaction with its environment and hence its response to changes in environmental conditions and management (Battaglia and Sands, 1998). In addition, complex physiological relationships can be replaced with simpler and lower resolution (and sometimes more empirical) relationships. This can reduce data input and parameterisation requirements and provide outputs at the desired resolution without any need to scale up. In some cases, the hybrid models provide even better growth predictions than empirical models developed for the same regions, probably partly because they account for intra- and inter-annual climatic variability (Weiskittel et al., 2010, Pérez-Cruzado et al., 2011).

The forest growth model 3-PG (Physiological Principles Predicting Growth) is an example of such a model. It is a simple process-based (or hybrid) model that has been widely used and well validated for all the processes it considers, such as growth, biomass partitioning, light absorption, water balance, responses to CO2, diameter distributions and mortality (Landsberg and Sands, 2010). It is widely used in South America as a management tool to complement routine inventory and strategic planning (Landsberg, 2003, Almeida et al., 2004b, Battaglia et al., 2007). It has been incorporated into a decision support system to assess plantation water yield and productivity in South Africa (Dye, 2005) and into the FullCam model for carbon accounting of Australian forests (Paul et al., 2006). Many more examples are described in Landsberg and Sands (2010). This wide-spread use is unusual for a process-based model and reflects its simple and transparent structure, the feasible input requirements, the relatively easily testable outputs for each process that it models, and importantly, the fact that it has always been freely available and well documented (Landsberg and Sands, 2010).

The 3-PG model was developed for even-aged monocultures and the aim of this study was to modify 3-PG so that it could be used to examine the growth dynamics of each species within a mixed-species forest and their responses to disturbances that modify stand density, such as thinning. The 3-PG model was chosen because no other models with comparable simplicity could be found that have been so rigorously tested in terms of observed-predicted comparisons for all of their component processes in so many different types of forests. In addition, to reproduce the spatial and temporal dynamics of species interactions in mixtures, models probably need to account for the physiological functioning of the trees, the environmental conditions within the stand, the stand structure, and interactions between these, as well as the external influences of climate and disturbances (Pretzsch et al., 2015). These components are all included in 3-PG or could be added with minor modifications. This was done for subtropical forests containing Castanopsis sclerophylla, Cunninghamia lanceolata, Cyclobalanopsis glauca and Liquidambar formosana, with a wide range of stand densities in Shitai County, Anhui Province, China. Mixed-species plantations of the same or related species have also been considered in Southern China (Meng et al., 2014). Specifically, the objectives were (1) to convert 3-PG into a mixed-species model and parameterise it for these species, one of which is deciduous; (2) to calibrate it using monospecific stands and assess its performance when predicting mixed-species dynamics and (3) to examine the spatial and temporal dynamics of the interactions between these species.

Section snippets

Description of 3-PG

3-PG was developed by Landsberg and Waring (1997). A review of the 3-PG model, its calibration and validation and many examples of its application is in Landsberg and Sands (2010). 3-PGPJS 2.7 (Sands, 2010) was used to develop the mixed-species version of 3-PG, 3-PGmix, which can still be run as 3-PGPJS 2.7. The 3-PGmix model is described in detail in Appendix A. The following provides an overview of the model followed by the changes that were required for it to work in mixed-species forests,

Changes made to produce 3-PGmix

The modifications described below were made to 3-PGPJS 2.7, which is freely available as an Excel file at http://3pg.forestry.ubc.ca/. The 3-PGmix model is also freely available as an Excel file at the same web site.

Site description

Model predictions were compared with data collected from subtropical monospecific Cunninghamia lanceolata plantations that were about 14- to 20-years-old and monospecific and mixed-species forests containing C. sclerophylla, C. lanceolata and L. formosana that were about 16- to 42-years-old and located in Shitai County, Anhui Province, China (29°59′–30°24′N, 117°12′–117°59′E). The stands varied in terms of L from <1 to 10.2, trees per ha from 88 to 2741 and basal area from 1 to 57 m2 ha−1. In

Results

Predicted and observed comparisons were made for monocultures using the light sub-model of 3-PGPJS 2.7 (L1) and the new light sub-model (L2) (Table 2). The comparisons were very similar and the new light model did not offer any significant improvement or result in any significant deterioration of predictions for these monospecific stands. Comparisons for mixtures could not be made because L1 cannot be used to predict the light absorption of mixtures.

The 3-PGmix model produced accurate

Discussion

The stands used for this study included a wide range of stand structures, in terms of stand density, species composition and species proportions as well as soil fertility. The 3-PGmix model was able to predict the growth, biomass partitioning and light absorption of the individual species within these mixed-species forests. The simulations also indicated that the interactions between these species (and complementarity effects) varied with stand age, stand density, fertility and rainfall. The

Acknowledgements

This study was part of a Sino-German Cooperation on Innovative Technologies and Service Capacities of Multifunctional Forest Management project (Lin2Value, projec number 033L049-CAFYBB2012013) supported by the Federal Ministry of Education and Research (BMBF) and Chinese Academy of Forestry. We thank Dr. Lutz Fehrmann, Prof. Christoph Kleinn, Prof. Christian Ammer, Dr. Hans Fuchs, Sabine Schreiner, Dr. Haijun Yang, Dr. Torsten Vor, Rubén Guisasola and Dengkui Mo for assistance with the plot

References (108)

  • G.-S. Chen et al.

    Carbon storage in a chronosequence of Chinese fir plantations in southern China

    Forest Ecol. Manag.

    (2013)
  • S. Condés et al.

    Mixing effect on volume growth of Fagus sylvatica and Pinus sylvestris is modulated by stand density

    Forest Ecol. Manag.

    (2013)
  • R.A. Duursma et al.

    Bias in the mean tree model as a consequence of Jensen's inequality

    Forest Ecol. Manag.

    (2003)
  • L.J. Esprey et al.

    Understanding 3-PG using a sensitivity analysis

    Forest Ecol. Manag.

    (2004)
  • P.M. Feikema et al.

    Validation of plantation transpiration in south-eastern Australia estimated using the 3PG+ forest growth model

    Forest Ecol. Manag.

    (2010)
  • D.I. Forrester

    The spatial and temporal dynamics of species interactions in mixed-species forests: From pattern to process

    Forest Ecol. Manag.

    (2014)
  • D.I. Forrester

    A stand-level light interception model for horizontally and vertically heterogeneous canopies

    Ecol. Model.

    (2014)
  • D.I. Forrester et al.

    Interactive effects of simultaneously applied thinning, pruning and fertiliser application treatments on growth, biomass production and crown architecture in a young Eucalyptus nitens plantation

    Forest Ecol. Manag.

    (2012)
  • C.A. Gonzalez-Benecke et al.

    Parameterization of the 3-PG model for Pinus elliottii stands using alternative methods to estimate fertility rating, biomass partitioning and canopy closure

    Forest Ecol. Manag.

    (2014)
  • V. Grimm

    Ten years of individual-based modelling in ecology: what have we learned and what could we learn in the future?

    Ecol. Model.

    (1999)
  • R. Guisasola et al.

    Intra- and inter-specific differences in crown architecture in Chinese subtropical mixed-species forests

    Forest Ecol. Manag.

    (2015)
  • S. Härkönen et al.

    Estimating annual GPP, NPP and stem growth in Finland using summary models

    Forest Ecol. Manag.

    (2010)
  • Y. He et al.

    Carbon storage capacity of monoculture and mixed-species plantations in subtropical China

    Forest Ecol. Manag.

    (2013)
  • Y. Huang et al.

    Seasonal changes in Cyclobalanopsis glauca transpiration and canopy stomatal conductance and their dependence on subterranean water and climatic factors in rocky karst terrain

    J. Hydrol.

    (2011)
  • P.H.M. Janssen et al.

    Calibration of process-oriented models

    Ecol. Model.

    (1995)
  • M.C. Kirnbauer et al.

    Estimating the stormwater attenuation benefits derived from planting four monoculture species of deciduous trees on vacant and underutilized urban land parcels

    Urban For. Urban Green.

    (2013)
  • J.J. Landsberg et al.

    A generalised model of forest productivity using simplified concepts of radiation-use efficiency, carbon balance and partitioning

    Forest Ecol. Manag.

    (1997)
  • J.J. Landsberg et al.

    Performance of the forest productivity model 3-PG applied to a wide range of forest types

    Forest Ecol. Manag.

    (2003)
  • G. le Maire et al.

    Tree and stand light use efficiencies over a full rotation of single- and mixed-species Eucalyptus grandis and Acacia mangium plantations

    Forest Ecol. Manag.

    (2013)
  • P. Miehle et al.

    A comparison of four process-based models and a statistical regression model to predict growth of Eucalyptus globulus plantations

    Ecol. Model.

    (2009)
  • A.C. Oishi et al.

    Estimating components of forest evapotranspiration: A footprint approach for scaling sap flux measurements

    Agric. Forest Meteorol.

    (2008)
  • C. Peng

    Growth and yield models for uneven-aged stands: past, present and future

    Forest Ecol. Manag.

    (2000)
  • C. Pérez-Cruzado et al.

    Combining empirical models and the process-based model 3-PG to predict Eucalyptus nitens plantations growth in Spain

    Forest Ecol. Manag.

    (2011)
  • H. Pretzsch et al.

    Representation of species mixing in forest growth models. A review and perspective

    Ecol. Model.

    (2015)
  • M.G. Ryan et al.

    Age-related decline in forest productivity: pattern and process

    Adv. Ecol. Res.

    (1997)
  • P.J. Sands et al.

    Parameterisation of 3-PG for plantation grown Eucalyptus globulus

    Forest Ecol. Manag.

    (2002)
  • J.V. Soares et al.

    Modeling the water balance and soil water fluxes in a fast-growing Eucalyptus plantation in Brazil

    J. Hydrol.

    (2001)
  • J.L. Stape et al.

    A twin-plot approach to determine nutrient limitation and potential productivity in Eucalyptus plantations at landscape scales in Brazil

    Forest Ecol. Manag.

    (2006)
  • J.K. Vanclay et al.

    Evaluating forest growth models

    Ecol. Model.

    (1997)
  • W. Wang et al.

    Development of TRIPLEX-Management model for simulating the response of forest growth to pre-commercial thinning

    Ecol. Model.

    (2011)
  • L. Wei et al.

    3-PG simulations of young ponderosa pine plantations under varied management intensity: Why do they grow so differently?

    Forest Ecol. Manag.

    (2014)
  • G. Xenakis et al.

    Sensitivity and uncertainty analysis from a coupled 3-PG and soil organic matter decomposition model

    Ecol. Model.

    (2008)
  • Almeida, A.C., Sands, P.J., 2015. Improving the ability of 3-PG to model the water balance of forest plantations in...
  • A.C. Almeida et al.

    Use of a spatial process-based model to quantify forest plantation productivity and water use efficiency under climate change scenarios

    Interfacing Modelling and Simulation with Mathematical and Computational Sciences

    (2009)
  • M.M. Amoroso et al.

    Comparing productivity of pure and mixed Douglas-fir and western hemlock plantations in the Pacific Northwest

    Can. J. Forest Res.

    (2006)
  • F. André et al.

    Precipitation water storage capacity in a temperate mixed oak-beech canopy

    Hydrol. Process.

    (2008)
  • O. Andrén et al.

    ICBM: the introductory carbon balance model for exploration of soil carbon balance

    Ecol. Appl.

    (1997)
  • M. Battaglia et al.

    Process-based model in Eucalyptus plantation management: reality and perspectives

    Boletín del CIDEU

    (2007)
  • H. Bugmann

    A review of forest gap models

    Clim. Change

    (2001)
  • M.G.R. Cannell et al.

    Radiation interception and productivity of willow

    J. Appl. Ecol.

    (1987)
  • Cited by (84)

    View all citing articles on Scopus
    View full text