Periodic Reporting for period 1 - DivMass (The role of Diversity on tropical forest bioMass dynamics: effects of disturbance and biogeography)
Période du rapport: 2023-09-01 au 2025-08-31
The relationship between biodiversity and ecosystem functions (BEF) is a central yet debated topic in ecological research, as consistent empirical support remains contraditory. Diversity metrics are assumed to represent the amount of niche space filled due to the better partitioning of resource use between different species, thereby leading to higher overall ecosystem function. In attempts to elucidate the establishment of a causal link between biodiversity and productivity in hyper-diverse tropical forests, canopy packing (greater crown complementarity resulting in more densely packed canopies), seems to be one important yet understudied missing key. Adding to this complexity, recent findings suggest that BEF relationships are not universal, but rather context-dependent. In particular, disturbance history has been shown to disrupt the effect of biodiversity on productivity.
The project main objective was to investigate whether canopy packing mediates the diversity-productivity relationships in old-growth and disturbed forest stands.
A first objective of the project aimed to provide an understanding of how diversity drives variation in biomass and productivity through different ecological mechanisms acting on specific components of biomass. A second objective of the project aimed to investigate whether historical disturbances influence the current effect of diversity on biomass and biomass productivity. We expected canopy packing to be greater in tree communities composed of species that are dissimilar in terms of resource-use strategies, that spatially and functionally complementary canopies would harvest more light and use it more efficiently, leading to increased biomass gain, and that stronger canopy packing effects should be found in old-growth forests, and weaker effects in disturbed forests.
The project main objective was to investigate whether canopy packing mediates the diversity-productivity relationships in old-growth and disturbed forest stands.
A first objective of the project aimed to provide an understanding of how diversity drives variation in biomass and productivity through different ecological mechanisms acting on specific components of biomass. A second objective of the project aimed to investigate whether historical disturbances influence the current effect of diversity on biomass and biomass productivity. We expected canopy packing to be greater in tree communities composed of species that are dissimilar in terms of resource-use strategies, that spatially and functionally complementary canopies would harvest more light and use it more efficiently, leading to increased biomass gain, and that stronger canopy packing effects should be found in old-growth forests, and weaker effects in disturbed forests.
The project used forest inventory data from 12 6.25-ha permanent plots that belong to an experimental site at the Paracou Tropical Forest Research Station in French Guiana. The plots were established from 1984 to 1988 and are early monitored ever since. From the 12 plots, nine were subjected to different types and intensities of sylvicultural treatments and three plots were left untouched as controls.
In this first overarching objective of the project, the researcher developed a data analysis framework aimed to deal with the complexity of the questions and data while ensuring a transparent set of ecological assumptions. We argue that diversity could impact biomass through different ecological mechanisms involving different biomass components. Because these effects may add up to each other, their respective role in biodiversity-ecosystem function (BEF) relationships are often challenging to infer due to confounding effects. Here, disaggregating biomass and its components (number of trees and/or their mean above-ground volume and wood density) allows the comprehension of the dissociating independent diversity effect, such as the tree packing and the sampling effects, which are expected to act through different biomass components or set of components.
Because the positive effect of diversity on canopy packing is often invoked as a mechanism that explains positive diversity–productivity relationships in forests, we assumed that canopy packing can increase the number of trees and/or their mean above-ground volume, with no causal effect of canopy packing on the wood density component of AGB, which is instead related to the sampling effect and was thus not considered in our models. The effect of functional diversity on canopy structural diversity (represented by the lidar derived index Shannon evenness of Plant Area Density - PAD) is thus expected on the number of trees and on above-ground volume.
We considered the following fixed hypothesized paths: functional dispersion (multivariate and trait-by-trait) was used as a predictor of Shannon PAD evenness, which was considered as a predictor of biomass gain.
For calculation of functional dispersion, we considered several species traits related to shade tolerance, regeneration strategy, competitive ability, growth, and stature. These included leaf area, specific leaf area, leaf thickness, maximum stem diameter, and wood density. Trait data were obtained from previous field campaigns conducted in French Guiana.
In this first overarching objective of the project, the researcher developed a data analysis framework aimed to deal with the complexity of the questions and data while ensuring a transparent set of ecological assumptions. We argue that diversity could impact biomass through different ecological mechanisms involving different biomass components. Because these effects may add up to each other, their respective role in biodiversity-ecosystem function (BEF) relationships are often challenging to infer due to confounding effects. Here, disaggregating biomass and its components (number of trees and/or their mean above-ground volume and wood density) allows the comprehension of the dissociating independent diversity effect, such as the tree packing and the sampling effects, which are expected to act through different biomass components or set of components.
Because the positive effect of diversity on canopy packing is often invoked as a mechanism that explains positive diversity–productivity relationships in forests, we assumed that canopy packing can increase the number of trees and/or their mean above-ground volume, with no causal effect of canopy packing on the wood density component of AGB, which is instead related to the sampling effect and was thus not considered in our models. The effect of functional diversity on canopy structural diversity (represented by the lidar derived index Shannon evenness of Plant Area Density - PAD) is thus expected on the number of trees and on above-ground volume.
We considered the following fixed hypothesized paths: functional dispersion (multivariate and trait-by-trait) was used as a predictor of Shannon PAD evenness, which was considered as a predictor of biomass gain.
For calculation of functional dispersion, we considered several species traits related to shade tolerance, regeneration strategy, competitive ability, growth, and stature. These included leaf area, specific leaf area, leaf thickness, maximum stem diameter, and wood density. Trait data were obtained from previous field campaigns conducted in French Guiana.
In the large-scale logging experimental site in Paracou, French Guiana we showed that functional dispersion, particularly based on traits related to shade tolerance, increased biomass gains through canopy packing in undisturbed Amazonian forests. However, this effect was absent in previously logged stands, where forest structural diversity did not fully recover even ca. 40 years after logging activities subsided, a pattern observed in previous studies, but reported here for the first time in the Amazon. Our findings indicate that the loss of the largest trees from the logging activities drastically changed the ecological process driving biomass productivity, altering light partitioning among species and disrupting the functional link between species composition, canopy packing, and biomass productivity.
To achieve objects 1, our results showed that undisturbed forests with higher multivariate functional dispersion showed higher Shannon PAD evenness (more homogeneous vertical distribution of plant material). Therefore, our findings confirm that higher canopy packing is achieved in communities characterized by higher functional complementarity (represented in this study by functional dispersion), where shade-tolerant co-exist with light-demanding and pioneer species, fostering the formation of functionally diverse tree species communities in terms of physiological adaptation to shade. Following our causal conceptual model, we found that an increase in Shannon PAD evenness led to an increase in biomass gain. More homogeneous occupancy of the canopy space is related to improved niche partitioning due to more efficient allocation of space and light resources. In older forest stands, the gaps created by large trees falling are more frequent, allowing the existence of niches that can be filled again by small trees . Those small trees do not directly compete with the largest trees due to vertical and horizontal stratification and their different sizes and shapes, but they can contribute to additional productivity to the whole stand due to a more complete use of resources. Indeed, the canopy packing effect was strongly related to tree recruitment. These findings indicate that niche complementarity (possibly related to gap dynamics) is an important determinant of forest productivity.
To achieve object 2, we found that the effect of functional dispersion on canopy packing is context-dependent, with functional dispersion not increasing Shannon PAD evenness in disturbed plots. In disturbed plots, Shannon PAD evenness showed no significant effect on biomass gain.
The project results show that the relationship between diversity, canopy packing, and productivity is strongly context dependent, and that, while enhancing biodiversity or canopy complexity may be beneficial in some circumstances, generalizations of nature-based initiatives—especially those aiming to boost biomass productivity—can be misleading, as biomass dynamics in tropical forests are governed by multiple interacting processes and mostly by the different land-use histories of secondary forests.
To achieve objects 1, our results showed that undisturbed forests with higher multivariate functional dispersion showed higher Shannon PAD evenness (more homogeneous vertical distribution of plant material). Therefore, our findings confirm that higher canopy packing is achieved in communities characterized by higher functional complementarity (represented in this study by functional dispersion), where shade-tolerant co-exist with light-demanding and pioneer species, fostering the formation of functionally diverse tree species communities in terms of physiological adaptation to shade. Following our causal conceptual model, we found that an increase in Shannon PAD evenness led to an increase in biomass gain. More homogeneous occupancy of the canopy space is related to improved niche partitioning due to more efficient allocation of space and light resources. In older forest stands, the gaps created by large trees falling are more frequent, allowing the existence of niches that can be filled again by small trees . Those small trees do not directly compete with the largest trees due to vertical and horizontal stratification and their different sizes and shapes, but they can contribute to additional productivity to the whole stand due to a more complete use of resources. Indeed, the canopy packing effect was strongly related to tree recruitment. These findings indicate that niche complementarity (possibly related to gap dynamics) is an important determinant of forest productivity.
To achieve object 2, we found that the effect of functional dispersion on canopy packing is context-dependent, with functional dispersion not increasing Shannon PAD evenness in disturbed plots. In disturbed plots, Shannon PAD evenness showed no significant effect on biomass gain.
The project results show that the relationship between diversity, canopy packing, and productivity is strongly context dependent, and that, while enhancing biodiversity or canopy complexity may be beneficial in some circumstances, generalizations of nature-based initiatives—especially those aiming to boost biomass productivity—can be misleading, as biomass dynamics in tropical forests are governed by multiple interacting processes and mostly by the different land-use histories of secondary forests.