Diversity in Resource Use Strategies Promotes Productivity in Young Planted Tree Species Mixtures

Abstract

Mixed-species forestry is a promising approach to enhance productivity, increase carbon sequestration, and mitigate climate change. Diverse forests, composed of species with varying structures and functional trait profiles, may have higher functional and structural diversity, which are attributes relevant to a number of mechanisms that can influence productivity. However, it remains unclear whether the context-dependent roles of functional identity, functional diversity, and structural diversity can lead to a generalized understanding of tree diversity effects on stand productivity. To address these gaps, we analyzed growth data from 83,600 trees from 89 species across 21 young tree diversity experiments spanning five continents and three biomes. Results revealed a positive saturating relationship between tree species richness and stand productivity, with reduced variability in growth rates among more diverse stands. Structural equation modeling demonstrated that functional diversity mediated the positive effects of species richness on productivity. We additionally report a negative relationship between structural diversity and productivity, which decreased with increasing species richness. When partitioning net diversity effects, we found that selection effects played a dominant role in driving the overall increase in productivity in these predominantly young stands, contributing 77% of the net diversity effect. Selection effects increased with diversity in wood density. Furthermore, acquisitive species with lower wood density and higher leaf nitrogen content had higher productivity in more diverse stands, while conservative species showed neutral to slightly negative responses to species mixing. Together, these results suggest that combining acquisitive with conservative species allows acquisitive species to drive positive selection effects while conservative species tolerate competition. Thus, contrasting resource-use strategies can enhance productivity to optimize mixed-species forestry, with potential for both ecological and economic benefits.

Keywords: TreeDivNet; climate change mitigation; complementarity effects; forest management; functional traits; mixed‐species forest plantations; selection effects; tree species richness.

FIGURE 1

Locations of the 21 tree diversity experiments from the TreeDivNet network included in this study, colored by biome. Detailed information on each experimental site can be found in Table S1. Map lines delineate study areas and do not necessarily depict accepted national boundaries.

FIGURE 2

Relationship between stand productivity and species richness across 21 experiments and three biomes. Stand productivity corresponds to standardized annual basal area increment. Positive values indicate productivity levels above the average within each experiment. Absolute mean stand productivity (m² ha⁻¹ year⁻¹) can be found in Table S3. Colors indicate biome. The dashed black line represents the mean stand productivity within each experiment. The solid black line represents the estimated mean productivity, with the shaded area showing a 95% confidence interval.

FIGURE 3

SEM illustrating direct and indirect links between (a) species richness and standardized stand productivity across 16 experiments (b) and the modeled interaction between species richness and structural diversity on stand productivity. The model displays standardized path coefficients for each pathway, and marginal R ² values for each endogenous variable. Blue pathways indicate positive correlations, while orange pathways indicate negative correlations. Significant pathways (p < 0.05) are shown with solid lines. Individual pathways are modeled separately in Figure S3.

FIGURE 4

The relationship of selection (a, d; yellow), complementarity (b, e; green) and net diversity (c, f; blue) effects (m² ha⁻¹ year⁻¹) with community weighted means (CWMs; a–c) and functional diversity (FDis; d–f) of wood density (WD) across 20 experiments. CWMs were z‐score standardized and FDis values were min–max standardized prior to analysis. Solid lines show the fitted values for variables that had a significant relationship (dashed lines for non‐significant) (Tables S5–S7). The shaded areas represent a 95% confidence interval. Illustrations showing individual data points can be found in Figure S4.

FIGURE 5

Relationship between species‐specific productivity and species richness across a gradient of more acquisitive species (low wood density [WD] and high leaf Nitrogen content [LNC]) to more conservative species (high WD and low LNC) across 20 experiments. High and low WD and LNC refer to values that are one standard deviation above and below the mean, respectively. Species‐specific productivity corresponds to annual basal area increment, standardized and log‐transformed before analysis. Both WD and LNC were z‐score standardized across experiments. Fitted values were back‐transformed from a logarithmic scale prior to illustration, and the shaded areas show the 95% confidence interval for the fitted model. An illustration showing individual data points can be found in Figure S5.