Non-additive biotic interactions improve predictions of tropical tree growth and impact community size structure

Abstract

Growth in individual size or biomass is a key demographic component in population models, with wide-ranging applications from quantifying species performance across abiotic or biotic conditions to assessing landscape-level dynamics under global change. In forest ecology, the responses of tree growth to biotic interactions are widely held to be crucial for understanding forest diversity, function, and structure. To date, most studies on plant–plant interaction only examine the direct competitive or facilitative interactions between species pairs; however, there is increasing evidence of indirect, higher-order interactions (HOIs) impacting species demographic rates. When HOIs are present, the dynamics of a multi-species community cannot be fully understood or accurately predicted solely from pairwise outcomes because of how additional species “interfere” with the direct, pairwise interactions. Such HOIs should be particularly prevalent where species show nonlinear functional responses to resource availability and resource-acquisition traits themselves are density dependent. With this in mind, we used data from a tropical secondary forest—a system that fulfills both of these conditions—to build a ontogenetic diameter-growth model for individuals across ten woody-plant species. We allowed both direct and indirect interactions within communities to influence the species-specific growth parameters in a generalized Lotka–Volterra model. Specifically, indirect interactions entered the model as higher-order quadratic terms, i.e. non-additive effects of conspecific and heterospecific neighbour size on the focal individual’s growth. For the whole community and for four out of ten focal species, the model that included HOIs had more statistical support than the model that included only direct interactions, despite the former containing a far greater number of parameters. HOIs had comparable effect sizes to direct interactions, and tend to further reduce the diameter growth rates of most species beyond what direct interactions had already reduced. In a simulation of successional stand dynamics, we show that such a further reduction in diameter growth by HOIs is important in reducing size asymmetry—and potentially less competitive exclusion due to shading—in a highly light-competitive forest. Our study highlights the potential role of higher-order interactions in stabilising communities in diverse forests.

Environmental gradients modify biotic interactions among tree species in temperate rain forest

“…To examine the effects of both abiotic and biotic factors on tree diameter growth, we calculated the annual diameter growth rate, G mipq = ∆Dmipq ∆t (cm yr −1 ), of tree m of species i in plot p and subplot q (see below for an explanation of subplot) as the change in DBH, ∆D mipq (cm), over census interval, ∆t (yr). Following Condit et al (2017) and Lai et al (2021), we assume the generative process of G mipq to follow a normal distribution (Equation (??)). This accommodates the zero and negative growth rates that comprised about a quarter of our data; ignoring them could introduce a systematic bias in the estimation of abiotic and biotic effects on growth.…”

“…Conspecific neighbour basal areas, N ipq , were calculated without the basal area of conspecific focal individual m. The pairwise interaction coefficients, α ij , quantify the per-basal-area main effects of species j on the growth of focal individual m (of species i). As density dependence among tree species has been shown to be non-additive (Lai et al 2021;Li et al 2021) (Box 1), we also included non-additive biotic interaction terms, j, k≥j β ijk N jpq N kpq , where each parameter β ijk quantifies the moderating effect of the density of the kth intermediary neighbour species, N kpq , on the main effect of direct neighbour species j in the same subplot. By modifying the pairwise interaction α ij between focal species i and direct neighbour j, the parameters β ijk are referred to as the higher-order interaction effects of neighbour species k on focal species i (Mayfield & Stouffer 2017;Kleinhesselink et al 2019;Letten & Stouffer 2019) (Box 1).…”

“…We opted to include or exclude entire sets of variables that share biological meaning, rather than constructing numerous candidate models with all possible variable combinations to avoid data dredging. However, this forces us to treat the importance of abiotic and biotic factors, as well as their interactions, as an 'all or none' question and assume that their effects are either equally important or equally unimportant (Lai et al 2021;Martyn et al 2021). Because there is no obvious biological basis for that assumption, we used Bayesian shrinkage simultaneously to achieve parsimony.…”

“…The key difference between non-additivity due to nonlinear density dependence and interaction modifiers is that the former process does not require the presence of a third species (for detailed explanation see Kleinhesselink et al 2019). Regardless of their mechanisms, accounting for non-additive biotic interactions has been shown to improve the prediction of tree growth in natural forest ecosystems (Lai et al 2021;Li et al 2021).…”

“…An inability to do so will prevent us from correctly identifying the drivers of community patterns and the relative importance of abiotic and biotic drivers under different circumstances (Freckleton et al 2009;Mayfield & Levine 2010;Cadotte & Tucker 2017). There have been empirical efforts to detect indirect abiotic effects (Coomes & Allen 2007;Kunstler et al 2011;Soliveres et al 2015;Bimler et al 2018) or indirect biotic effects (Uriarte et al 2004;Mayfield & Stouffer 2017;Xiao et al 2020;Lai et al 2021;Li et al 2021) in plant-plant interactions, but rarely both.…”