Tropical tree allometry and crown allocation, and their relationship with species traits in central Africa
Introduction
Accurate assessments of biomass and carbon stocks in tropical forests underpin policies that aim to mitigate carbon dioxide emissions such as the UN-REDD + program and the recommendations of the Intergovernmental Panel on Climate Change (Gibbs et al., 2007) and using allometric equations to predict the aboveground biomass (AGB) of tropical trees is a cost-effective and accurate approach (Chave et al., 2004, 2014). In the last decades, general multi-specific allometric models have been updated, with the pantropical model of Chave et al. (2014) refining the previously developed models (Brown et al., 1989, Chave et al., 2005). The most recent pantropical model of Chave et al. (2014) has been shown to perform well at the tree level, on independent dataset of destructive AGB in a peat swamp forest in Indonesia (Manuri et al., 2014) and in six terra firme forest sites distributed across central Africa (Fayolle et al., 2018).
After tree diameter, species wood density and total tree height are both important AGB predictors for tropical trees (Brown et al., 1989, Chave et al., 2014, Chave et al., 2005, Feldpausch et al., 2011, Van Breugel et al., 2011), respectively expected to explain between-species and between-site variations (Ketterings et al., 2001). The inclusion of height has indeed been shown to offset allometric variation across forest types (Chave et al., 2005), and when not available locally, height can possibly be predicted from tree diameter using a local or a regional equation (Feldpausch et al., 2011) or using a general equation including an environmental stress variable (Chave et al., 2014). In addition to the classical AGB predictors, the use of crown information has recently received attention (Goodman et al., 2014, Fayolle et al., 2018), and its importance to predict AGB has been notably explained by ontogenic and/or size-related changes in biomass allocation to crown (Ploton et al., 2016). Besides wood density averaged at species (Molto et al., 2013) or even at genus level (Slik, 2005), interspecific variation in tropical tree allometry and allocation have not been deeply explored. Because it is extremely costly, time-consuming, and difficult to collect and assemble biomass data at the tree and species levels (Picard et al., 2012), many questions about the usefulness of generic models remain unanswered. Still, it is unclear to what extent data on biomass allometry should be pooled or separated according to morphological, phylogenetic and/or phenological characteristics of species (Paul et al., 2016). In the latter study, specific allometric equations have been developed for species grouped into a few plant functional types covering contrasted biomes and ecoregions across Australia. It has also been argued that species-specific models are especially needed when the gains in accuracy at the stand level are high, which is true for high-value monocultures (Paul et al., 2016) or for monodominant species such as Gilbertiodendron dewevrei in central Africa, and for which a specific allometry has been developed (Umunay et al., 2017). Specific allometric relationships have been developed for species grouped at genus level in a Dipterocarp forest of south-eastern Asia (Basuki et al., 2009) or by family and wood density in peat swamp forests in Indonesia (Manuri et al., 2014). In the latter study, it has to be noted, however, that the multi-specific pantropical model showed similar performance as the local models.
Interspecific variation in tropical tree allometry has been demonstrated in the early stages (seedlings and saplings) and related to life-history strategy and traits in Panama (King, 1990, King, 1996). The allometric specialization of light-demanding species was found to be associated with biomass allocation toward efficient height growth, and that of understory species with the maximization of light interception and the persistence in the shaded understory (King, 1990). Light-demanding species are indeed generally more exposed to light at 10 cm diameter (Sheil et al., 2006) because of greater height, as reported in Bolivia (Poorter et al., 2006) and in northern Congo (Loubota Panzou et al., 2018). It is thus important to both consider allometry and allocation in evaluating the adaptive significance of interspecific variation. In addition, grouping tropical tree species into meaningful groups of tree species is not straightforward (Swaine and Whitmore, 1988) and only a few plant functional types are generally recognized in Dynamic Global Vegetation Models (DGVM, Fisher et al., 2018), generally two groups, evergreen vs deciduous tropical trees. In the successional model TROLL (Maréchaux and Chave, 2017), up to 12 tropical plant functional types are recognized, and for each of them, the tree geometry is modelled explicitly with specific allometric relationships relating tree diameter, to height, crown radius and depth. Though plant functional types are at the core of many models of vegetation dynamics or of forest succession, the lack of information on allometry, allocation and traits, for some species and regions hampers models predictions and this is particularly true for central Africa which is a largely under-sampled region.
In this study, we evaluated how size-dependent changes in above-ground biomass and biomass allocation to crown relate to other allometric and life-history traits for tropical tree species. We addressed three questions. First, to what extent biomass allometry and allocation vary among a set of tropical tree species? Using published data of destructive biomass in central Africa, we were able to provide new information on biomass allometry and allocation for 54 tropical tree species widespread and/or locally abundant, and covering a variety of genera and families. Second, is interspecific variation in allometry and allocation related to key traits? We notably tested the hypothesis of contrasted biomass allometry and allocation, between evergreen and deciduous species, but besides these two groups, we further investigated interspecific variation in response to life-history traits extracted from literature and to allometric traits newly developed for 50 species, and characterizing species architecture, light requirement, the wood economics spectrum, and adaptation to fire (Table 1). Third, which allometric or life-history traits can be used in multi-specific allometry and allocation models as a surrogate of interspecific variation? We notably tested the hypothesis that wood density explains interspecific variation in biomass allometry (Ketterings et al., 2001) and that light requirement explains interspecific variation in biomass allocation to crown (King, 1990).
Section snippets
Study sites
Destructive biomass data were available for more than 1,000 trees sampled across central Africa in eight sites (Fig. 1) which includes the six sites sampled in the PREREDD+ project (sites #1–2 and #5–8 on the map, Fayolle et al., 2018), and two sites sampled earlier, the Zadié site in Gabon (#3, Ngomanda et al., 2014) and the Mindourou site in Cameroon (#4, Fayolle et al., 2013). In each site, the sampling covered a large number of trees (≥100) and a vast range of diameters and wood density, as
Tree allometry
Benefitting from existing destructive AGB data in eight sites across central Africa, we were able to fit species-specific allometric models for 54 tropical tree species (5–69 trees per species, mean of 19). As expected, we found strong interspecific variation in tree allometry when only considering the bivariate relationships between tree AGB and D (Fig. 2). The coefficients of AGB models varied among the 54 tropical tree species with a mean of 2.45 (± 0.24 standard deviation) found for the
Discussion
In accordance with the predictions of the metabolic theory, under certain hypotheses of biomechanical constraints of tree stability and hydraulic resistance in conductive cell networks, tree AGB scales with tree diameter with an exponent of 8/3 (West et al., 1997). The metabolic theory has been hotly debated, specifically its general nature and its underlying assumption of invariant scaling coefficient (Zianis and Mencuccini, 2004, Muller-Landau et al., 2006). Here, the scaling coefficient of
Conclusions
Using published destructive biomass data of tropical trees in eight sites representative of terra firme forests across central Africa we were able to provide new information on biomass allometry and allocation, for a set of 54 tropical tree species widespread and/or locally abundant, and phylogenetically dispersed. New allometric traits were derived for 50 species and were both related to the coefficients of the allometry and allocation models, and were also included in multi-specific models in
CRediT authorship contribution statement
Géraud Sidoine Mankou: Conceptualization, Formal analysis, Investigation, Writing - original draft, Writing - review & editing. Gauthier Ligot: Conceptualization, Formal analysis, Investigation, Writing - review & editing. Grace Jopaul Loubota Panzou: Conceptualization, Investigation, Writing - review & editing. Faustin Boyemba: Data curation, Funding acquisition, Project administration, Writing - review & editing. Jean-Joël Loumeto: Data curation, Funding acquisition, Project administration,
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
G.M. benefited from a fellowship of the International Tropical Timber Organization [ITTO grant #084/17A]. The data collection was supported by Gembloux Agro-Bio Tech, Université de Liège through the EBALAC project, by the ACP Secretariat and the European Commission under the EUACP ‘‘Establishment of a forestry research network for ACP countries’’ [9 ACP RPR 91#1–FORENET], and by a gift from the Global Environment Fund [TF010038] administered by the World Bank to the COMIFAC which funded the
References (53)
- et al.
Allometric equations for estimating the above-ground biomass in tropical lowland< i> Dipterocarp</i> forests
For. Ecol. Manag.
(2009) - et al.
Tree allometry in Central Africa: Testing the validity of pantropical multi-species allometric equations for estimating biomass and carbon stocks
For. Ecol. Manag.
(2013) - et al.
Taller trees, denser stands and greater biomass in semi-deciduous than in evergreen lowland central African forests
For. Ecol. Manag.
(2016) - et al.
Reducing uncertainty in the use of allometric biomass equations for predicting above-ground tree biomass in mixed secondary forests
For. Ecol. Manag.
(2001) - et al.
Site-specific versus pantropical allometric equations: Which option to estimate the biomass of a moist central African forest?
For. Ecol. Manag.
(2014) Assessing tropical lowland forest disturbance using plant morphological and ecological attributes
For. Ecol. Manag.
(2005)- et al.
Estimating biomass and carbon for Gilbertiodendron dewevrei (De Wild) Leonard, a dominant canopy tree of African tropical Rainforest: Implications for policies on carbon sequestration
For. Ecol. Manag.
(2017) - et al.
On simplifying allometric analyses of forest biomass
For. Ecol. Manag.
(2004) - et al.
Biomass estimation methods for tropical forests with applications to forest inventory data
For. Sci.
(1989) - et al.
Tree allometry and improved estimation of carbon stocks and balance in tropical forests
Oecologia
(2005)