Introduction
Mixed species plantations with high value species attract increased interest as they may provide a broader supply of ecological and socio-economic benefits (Felton et al., 2016; Gamfeldt et al., 2013; Heinrichs et al., 2019; Isbell et al., 2011). Mixed species stands have particularly been reported to provide higher biomass production (Erskine, Lamb, & Bristow, 2006) and to be more productive, stable and climate resistant than the average monocultures of the same species (Bauhus et al., 2017; H. Pretzsch et al., 2015; H Pretzsch et al., 2019).
As compared to monoculture, higher biomass production in mixed species stands can be attributed to several factors, including greater species diversity, greater stand structuring and canopy packing (H Pretzsch et al., 2019), facilitation and better resource utilization (Jactel et al., 2018; Hans Pretzsch, Forrester, & Bauhus, 2017). For instance, species rich stands may allow coexistence of functionally different species (species of different functional trait or attributes) that therefore efficiently access and utilize limiting resources, thereby enhancing biomass production through efficient resource-use. It is further possible that in mixed species stands, certain species (e.g. nitrogen-fixing species) improve growing conditions for others, thereby enhancing overall production through facilitation (Erskine et al., 2006). This is consistent with the niche complementarity and facilitation mechanisms (Loreau & Hector, 2001), and have important implications for silviculture, as well managed mixed, more diverse and uneven-aged plantations would have higher net primary production than monoculture stand (Kelty, 1992, 2006). On the other hand, occurrence of dominant and highly productive tree species (with dominant traits) can positively influence biomass production, i.e. one or two species in mixed species plantation can largely explain increase in biomass production if they are dominant. This lends support to the sampling/selection effect hypothesis which posits that biomass production is enhanced through functional traits of the dominant species. Although the selection effects may seem more evident, because few larger trees often contain large portion of the stand aboveground biomass (Bastin et al., 2015; Fotis et al., 2018; Lin et al., 2016; S. Mensah, Veldtman, Du Toit, Kakaï, & Seifert, 2016; Sylvanus Mensah, Veldtman, & Seifert, 2017), studies have also lent support to both mechanisms, which are demonstrated to be non-mutually exclusive (Cavanaugh et al., 2014; Hooper, Chapin III, & Ewel, 2005; S. Mensah, du Toit, & Seifert, 2018; S. Mensah, Veldtman, Assogbadjo, Glèlè Kakaï, & Seifert, 2016; Ruiz-Benito et al., 2014; Ruiz-Jaen & Potvin, 2010; Wu et al., 2015), but can have different relative importance in different contexts (Fargione et al., 2007; Potvin & Gotelli, 2008), as a result of differences in functional traits among species, resource allocation and resource use efficiency (Huston, 1997; Tilman, Lheman, & Thomson, 1997). For instance in a recent study, we showed that both mechanisms operate through competitive exclusion imposed by dominant species (selection effects) and complementary use of resources by weak competitors (S. Mensah, du Toit, et al., 2018). Therefore both mechanisms may also prevail for mixed species stands, and understanding their relative contribution may inform about appropriate silvicultural options for their management.
Decades of research have helped establish positive relationships between species richness and ecosystem biomass or carbon storage, as the most dominant pattern. Across scales and biomes, several studies have reported positive effects of species diversity on stand biomass (Barrufol et al., 2013; Cheng, Zhang, Zhao, & von Gadow, 2018; Huang, Su, Li, Liu, & Lang, 2019; Liang et al., 2016; Liu et al., 2018; S. Mensah, Veldtman, Du Toit, et al., 2016; Paquette & Messier, 2011; Ruiz-Benito et al., 2014; Vilà et al., 2007), although neutral and negative patterns also exist (An-ning, Tian Zhen, & Jian Ping, 2008; Ruiz-Jaen & Potvin, 2011; Szwagrzyk & Gazda, 2007). While the positive relationship between species richness and biomass production can be used as a persuasive argument for the conservation of biodiversity and encourage more diverse plantations (Erskine et al., 2006), many previous studies have focused on species richness or related taxonomic indexes, which do not fully capture certain functional differences or similarities between species (Cardinale et al., 2006), nor are they sufficient to reflect the complexity of the stand community (Morin, Fahse, Scherer-Lorenzen, & Bugmann, 2011). Much research is still needed across scales of the analysis (global, national or subnational), and in relation to the measure of biodiversity.
Apart from species richness or related taxonomic indices such as Shannon index, Pielou evenness and Simpson index, functional trait diversity (richness, evenness, dispersion and divergence), functional trait dominance or identity (community-weighted mean of a given functional trait) or structural diversity have been reported to predict stand aboveground biomass or productivity (Y. Li et al., 2019; Lin et al., 2016; S. Mensah, du Toit, et al., 2018; S. Mensah, Veldtman, Assogbadjo, et al., 2016; S. Mensah, Veldtman, Du Toit, et al., 2016; Prado-junior et al., 2016; Thom & Keeton, 2019; Wen et al., 2019; Zhang & Chen, 2015). Because species may differ in functional traits that drive differences in resource capture, rates of photosynthesis and biomass allocation (Falster, Duursma, & FitzJohn, 2018; Poorter et al., 2012), functional trait diversity, dominance or identity would better capture the degree of functional redundancy and niche overlap (Lasky et al., 2014; Prado-junior et al., 2016; Ruiz-Benito et al., 2014). Further, structural diversity (tree size variation and inequality) reflects how different species occupy different vertical and horizontal layers, and therefore may indicate the degree of complementarity (e.g. light-adapted and shade-tolerant species), competition and resource utilisation. Nevertheless, some recent studies showed controversy in the relationship between these structural and functional diversity/dominance metrics (Finegan et al., 2015; Lin et al., 2016; Prado-junior et al., 2016; Xu et al., 2019). For instance, Lin et al. (2016), after accounting for topographic variables and tree stem density, found that functional dominance was the main driving factor for forest aboveground carbon, while functional diversity had negligible effects. The authors argued that this could have been due to the fact functional traits that relate strongly to plant complementary resource use were not included in the analysis. However, even after using five functional traits (maximum tree height; leaf carbon content; leaf nitrogen content; leaf area; and specific leaf area), Xu et al. (2019) reported that neither the functional nor the phylogenetic diversity showed a significant advantage in predicting aboveground biomass and biomass production when compared with species richness in old-growth temperate forests. Further, Fotis et al. (2018) recently reported limited effect of functional diversity on aboveground biomass in mixed mesophytic temperate forests of the eastern USA, whereas Szwagrzyk and Gazda (2007) found that negative effects of functional diversity on aboveground biomass in Central Europe. Consequently, the relative importance of functional dominance (selection effects), functional diversity and structural diversity (niche complementarity) for stand biomass is still controversial, and requires further investigation, especially in mixed species stands that harbor species occupying different positions across the vertical layer, possibly favoring a better use of resources (e.g., light) and reduced competition.
Combining information on taxonomic, functional, and structural diversity would provide additional insights into our understanding of mechanisms behind diversity-biomass relationships in mixed species stands. However, it is unclear how each particular metric would predict AGB, and whether some have significant advantage in mediating AGB response. Therefore, in this study, we used taxonomy-, structure-, and functional trait-based diversity to examine the relationships between AGB and multiple diversity metrics. First we sought to determine the most important taxonomic diversity measure among species richness, Shannon diversity, Pielou evenness and Simpson index, for predicting AGB. Second we tested for clear effects of multiple trait-functional diversity, single trait-functional diversity, functional dominance, and structural diversity on AGB. Finally, we retained the most important structure-, and functional trait-based diversity metrics, and tested for their mediation role in predicting AGB response to species richness.