L’agroforesterie, souvent présentée comme solution à l’érosion de la biodiversité et aux chocs climatiques, échoue fréquemment lorsqu’elle repose sur des programmes massifs de plantation. Une autre approche, centrée sur la parcelle, valorise la complémentarité entre arbres rémanents, spontanés et plantés. Cette complémentarité renforce biodiversité, carbone, productivité et durabilité des cacaoyères.

Secondary forest succession following agricultural abandonment is a dominant land-use transition across the tropics, yet whether soil microbial communities recover toward old-growth forest reference states remains poorly understood, particularly in West Africa. Here, we investigated the successional dynamics of bacterial and arbuscular mycorrhizal (AM) fungal communities along post-agricultural chronosequences spanning 1 to 43 years across six classified forests in Côte d'Ivoire, using Bayesian hierarchical models applied to amplicon sequencing data. Both guilds attained moderate to high alpha diversity within the first decade of succession; AM fungal diversity showed moderate evidence of age-related increase thereafter while bacterial diversity showed no directional trend. Pairwise turnover analyses revealed progressive internal convergence in AM fungal communities with plots farther apart in successional time becoming more compositionally similar, while bacterial communities showed only a weak and uncertain tendency in the same direction. Beta-dispersion modelling further indicated progressive within-forest homogenisation of AM fungal communities across abundance-weighted metrics, while bacterial assemblages showed no such stabilisation. Despite this internal convergence, compositional distances to old-growth reference plots remained persistently high for both guilds throughout the chronosequence, with no statistical evidence of recovery toward old-growth states across any dissimilarity metric or guild within the 40-year window. Indicator species analysis identified no robust stage-specific taxa after correction for multiple testing. These results indicate that microbial succession in post-agricultural West African forests is characterised by rapid early reorganisation followed by stabilisation into site-specific assemblages that remain persistently distinct from old-growth reference communities. This outcome challenges the direct application of classical vegetation successional theory to soil microbiomes and indicates that no detectable recovery toward old-growth microbial community states was observed within the 40-year chronosequence.

Cocoa cultivation in West Africa has been a major driver of deforestation, leading to increased greenhouse gas emissions and threatening cocoa yields. Agroforestry, which integrates trees from various origins-remnant, spontaneous, and planted-presents a sustainable solution to enhance carbon sequestration and improve farm resilience. However, the specific contributions of these tree origins and the socio-environmental factors shaping their effectiveness remain poorly understood. This study examines carbon dynamics across 150 cocoa fields in Côte d'Ivoire, analyzing a total of 11,568 trees across 15 sites. Using Bayesian modeling, we assess carbon stocks and gains from tree growth to explore how socio-environmental factors influence carbon balance in cocoa fields. Carbon stocks varied widely with remnant having the highest median carbon stocks (6.33 Mg/ha), followed by spontaneous (2.06 Mg/ha) and planted trees (1.53 Mg/ha). Carbon gains are similar for planted and spontaneous trees up to 7 years, but afterward, spontaneous trees grow faster (11.20 ± 0.87 kg/year) than planted ones (3.96 ± 0.5 kg/year). Carbon stocks rose with informed farmers and former forest use, but declined with higher cocoa density. Carbon gains at the tree level is primarily influenced by ownership and previous forest land use with positive effects, while cocoa density and annual temperature have negative effects. To maximize carbon sequestration and ensure the sustainable management of agroforestry systems, interventions should prioritize securing land tenure, enhancing farmer training in tree botany, and promoting the conservation of remnant and spontaneous trees.

Quantifying how soils resist degradation and recover organic carbon following agricultural abandonment is critical for understanding land restoration trajectories in West Africa. This study applies a Bayesian hierarchical modeling framework to assess resistance (the proportion of soil organic carbon [SOC] retained after land use) and resilience (the rate of SOC recovery) during forest succession across eight chronosequences in Cote d'Ivoire. SOC was measured across three depths (0-30 cm) under fallow, secondary, and old-growth forests. Modeling revealed that soil resistance was about 50 % of their original SOC, with lower resistance under cotton (42 %) and perennial crops (47 %), and higher under annual crops (57 %). Resistance also varied by topography, with lowerslope sites averaging 42 %, versus over 52 % at other positions. SOC resilience was highly influenced by environmental and management factors. Remnant tree density had the strongest positive effect: areas with over 50 remnant trees per hectare had recovery rates twice as high as sites without. Conversely, increasing precipitation, temperature, and their seasonality slowed SOC recovery, highlighting climate-related limits to resilience. SOC recovery also varied by soil type, with Acrisols showing the highest resilience. Overall, the model indicates that up to 90 % of SOC can be recovered within 10 years in favorable conditions. The developed Bayesian approach provides probabilistic estimates of SOC trajectories, offering nuanced insight into ecological and management drivers. Our findings emphasize the importance of preserving remnant trees as ecological anchors in postagricultural landscapes, reinforcing passive regeneration as a cost-effective strategy to restore soil carbon in West Africa.

Cocoa agroforestry systems (AFS) in West Africa represent an underexplored yet promising source of timber. Within these systems, species can reach a 50-cm diameter as early as 14 years of age. Naturally regenerated trees grow 10

Cocoa production stands as the primary driver of biodiversity loss in West Africa. In response to this challenge, agroforestry emerges as a promising solution to restore tree cover and preserve biodiversity in agricultural landscapes. Despite distributions of millions of trees to farmers through various sustainability initiatives, survival rates of planted trees have been disappointing in cocoa fields. However, many non-planted trees currently exist in cocoa fields, including trees from spontaneous regeneration and remnants from previous forests. Our study aims to explore the origins of tree biodiversity in current Ivorian cocoa fields and identify the key socio-environmental determinants. Across 150 cocoa fields, 9939 trees were botanically identified, allowing estimation of diversity and complementarity levels between planted, spontaneous, and remnant cohorts, linked to socio-environmental variables within a Bayesian modelling framework. Results revealed that remnant and spontaneous trees exhibited much greater diversity than planted trees, with the latter showing important complementarity with the other cohorts. Moreover, the diversity of each tree cohort was influenced by a distinct set of environmental factors such as climate and soil composition, highlighting the need to consider the unique history and location of each cocoa field to support agroforestry development. Effective biodiversity maximisation thus relies on preserving remnant trees, vital sources of propagules for natural regeneration, while carefully selecting spontaneous trees and complementary introducing planted trees. Furthermore, recognizing the farmer’s identity as a key factor shaping observed biodiversity underscores the importance of raising awareness among farmers and offering tree recognition training, particularly for young trees.

Ecology and forestry sciences are using an increasing amount of data to address a wide variety of technical and research questions at the local, continental and global scales. However, one type of data remains rare: fine-grain descriptions of large landscapes. Yet, this type of data could help address the scaling issues in ecology and could prove useful for testing forest management strategies and accurately predicting the dynamics of ecosystem services. Here we present three datasets describing three large European landscapes in France, Poland and Slovenia down to the tree level. Tree diameter, height and species data were generated combining field data, vegetation maps and airborne laser scanning (ALS) data following an area-based approach. Together, these landscapes cover more than 100 000 ha and consist of more than 42 million trees of 51 different species. Alongside the data, we provide here a simple method to produce high-resolution descriptions of large landscapes using increasingly available data: inventory and ALS data. We carried out an in-depth evaluation of our workflow including, among other analyses, a leave-one-out cross validation. Overall, the landscapes we generated are in good agreement with the landscapes they aim to reproduce. In the most favourable conditions, the root mean square error (RMSE) of stand basal area (BA) and mean quadratic diameter (Dg) predictions were respectively 5.4 m2.ha-1 and 3.9 cm, and the generated main species corresponded to the observed main species in 76.2% of cases.

Tropical forests play a crucial role in climate regulation due to their high carbon sequestration capacity. However, degradation and disturbances in these forests may result in significant carbon losses. This study focuses on the impact of various biophysical, anthropogenic, and landscape factors on aboveground biomass (AGB) in heavily disturbed landscapes of Côte d'Ivoire (West Africa), a typical low-forest and high-deforestation country. AGB estimates from a National Forest Inventory dataset have been linked to five categories of variables (Climate, Soil, Topography, Landscape, and Human-related) through a random forest modeling approach that addressed collinearity among variables, selected key variables from each category, and used spatial cross-validation to evaluate model performance. The comprehensive model, combining landscape composition, physical soil properties, and climate variables, demonstrated strong performance with an R-squared of 0.62. Notably, the percentage of landscape occupied by forest within a radius of 1000 m (PLAND1000) had a highly significant impact on AGB, exhibiting a notable increase when PLAND1000 exceeded 80 % and a decrease when it felt below 25 %. Soil properties, both physical (Bulk Density and Coarse Fraction) and chemical (soil pH), significantly influenced AGB, too. Interestingly, climatic, topographic, and other anthropogenic variables had minimal relevance in predicting AGB, suggesting that their effects may have been captured by landscape and soil integrative variables. In order to enhance forest preservation and restoration initiatives in the face of deforestation and fragmentation challenges in the West African region, we recommend (i) evaluating the appropriate landscape scale of effect (a 1000 m radius circle being the most significant in this study); (ii) prioritizing the preservation or restoration of dense forest landscapes; and (iii) integrating landscape composition into forest management policies.

Cocoa production has been one of the main drivers of forest loss in West Africa. In the resulting post-forest landscapes, agroforestry has often been recognised as a solution to reconcile the preservation of trees and agriculture. Thus, a large number of tree plantation programs have been carried out in cocoa fields. Despite these major investments, the success of these plantations as a tool for "reforestation" of landscapes and sustainable timber production has never really been evaluated in fields where remnant trees, spontaneous trees and (trans)planted trees coexist. To quantify the current and future timber resource, we inventoried all trees in 150 cocoa fields distributed along the bioclimatic and historical gradients of the cocoa production area of Côte d'Ivoire, the world's leading producer. Our results show that (i) 19.6% of all associated trees are timber species, (ii) in plots where farmers actually introduced trees by planting, only 13.1% of trees have been (trans)planted, (iii) 69.7% of the current timber volume comes from remnant trees and (iv) spontaneous trees constitute 77% of the future timber resource. Based on our results, we propose 23 species along with their cultivation methods for the renewal of timber resource in cocoa fields. Overall, our results show the failure of plantation programs in cocoa fields and suggest to bet on natural regeneration to sustainably provide timber wood. Consequently, private companies supplying trees to farmers should focus on species that are complementary to those already present in natural regeneration. At the landscape level, remnant trees and residual forests should be preserved to maintain propagule sources. Finally, investments in reforestation of cocoa fields should be redirected towards training small farmers in silvicultural management techniques such as assisted natural regeneration and tree pruning.

Ecology and forestry sciences are using an increasing amount of data to address a wide variety of technical and research questions at the local, continental and global scales. However, one type of data remains rare: fine-grain descriptions of large landscapes. Yet, this type of data could help address the scaling issues in ecology and could prove useful for testing forest management strategies and accurately predicting the dynamics of ecosystem services. Here we present three datasets describing three large European landscapes in France, Poland and Slovenia down to the tree level. Tree diameter, height and species data were generated combining field data, vegetation maps and airborne laser scanning (ALS) data following an area-based approach. Together, these landscapes cover more than 100 000 ha and consist of more than 42 million trees of 51 different species. Alongside the data, we provide here a simple method to produce high-resolution descriptions of large landscapes using increasingly available data: inventory and ALS data. We carried out an in-depth evaluation of our workflow including, among other analyses, a leave-one-out cross validation. Overall, the landscapes we generated are in good agreement with the landscapes they aim to reproduce. In the most favourable conditions, the root mean square error (RMSE) of stand basal area (BA) and mean quadratic diameter (Dg) predictions were respectively 5.4 m2.ha-1 and 3.9 cm, and the generated main species corresponded to the observed main species in 76.2% of cases.

Despite many studies showing biodiversity responses to warming, the generality of such responses across taxonomic groups remains unclear. Very few studies have tested for evidence of bryophyte community responses to warming, even though bryophytes are major contributors to diversity and functioning in many ecosystems. Here we report an empirical study comparing long-term change of bryophyte and vascular plant communities in two sites with contrasting long-term warming trends, using “legacy” botanical records as a baseline for comparison with contemporary resurveys. We hypothesized that ecological changes would be greater in sites with a stronger warming trend, and that vascular plant communities, with narrower climatic niches, would be more sensitive than bryophyte communities to climate warming. For each taxonomic group in each site, we quantified the magnitude of changes in species’ distributions along the elevation gradient, species richness, and community composition. We found contrasted temporal changes in bryophyte vs. vascular plant communities, which only partially supported the warming hypothesis. In the area with a stronger warming trend, we found a significant increase of local diversity and beta-diversity for vascular plants, but not for bryophytes. Presence absence data did not provide sufficient power to detect elevational shifts in species distributions. The patterns observed for bryophytes are in accordance with recent literature showing that local diversity can remain unchanged despite strong changes in composition. Regardless of whether one taxon is systematically more or less sensitive to environmental change than another, our results suggest that vascular plants cannot be used as a surrogate for bryophytes in terms of predicting the nature and magnitude of responses to warming. Thus, to assess overall biodiversity responses to global change, abundance data from different taxonomic groups and different community properties need to be synthesized.

Forest models are instrumental for understanding and projecting the impact of climate change on forests. A considerable number of forest models have been developed in the last decades. However, few systematic and comprehensive model comparisons have been performed in Europe that combine an evaluation of modelled carbon and water fluxes and forest structure. We evaluate 13 widely used, state-of-the-art, stand-scale forest models against field measurements of forest structure and eddy-covariance data of carbon and water fluxes over multiple decades across an environmental gradient at nine typical European forest stands. We test the models' performance in three dimensions: accuracy of local predictions (agreement of modelled and observed annual data), realism of environmental responses (agreement of modelled and observed responses of daily gross primary productivity to temperature, radiation and vapour pressure deficit) and general applicability (proportion of European tree species covered). We find that multiple models are available that excel according to our three dimensions of model performance. For the accuracy of local predictions, variables related to forest structure have lower random and systematic errors than annual carbon and water flux variables. Moreover, the multi-model ensemble mean provided overall more realistic daily productivity responses to environmental drivers across all sites than any single individual model. The general applicability of the models is high, as almost all models are currently able to cover Europe's common tree species. We show that forest models complement each other in their response to environmental drivers and that there are several cases in which individual models outperform the model ensemble. Our framework provides a first step to capturing essential differences between forest models that go beyond the most commonly used accuracy of predictions. Overall, this study provides a point of reference for future model work aimed at predicting climate impacts and supporting climate mitigation and adaptation measures in forests.

A growing body of research suggests mixed-species stands are generally more productive than pure stands as well as less sensitive to disturbances. However, these effects of mixture depend on species assemblages and environmental conditions. Here, we present the Salem simulator, a tool that can help forest managers assess the potential benefit of shifting from pure to mixed stands from a productivity perspective. Salem predicts the dynamics of pure and mixed even-aged stands and makes it possible to simulate management operations. Its purpose is to be a decision support tool for forest managers and stakeholders as well as for policy makers. It is also designed to conduct virtual experiments and help answer research questions. In Salem, we parameterised the growth in pure stand of 12 common tree species of Europe and we assessed the effect of mixture on species growth for 24 species pairs (made up of the 12 species mentioned above). Thus, Salem makes it possible to compare the productivity of 36 different pure and mixed stands depending on environmental conditions and user-defined management strategies. Salem is essentially based on the analysis of National Forest Inventory data. A major outcome of this analysis is that we found species mixture most often increases species growth, in particular at the poorest sites. Independently from the simulator, foresters and researchers can also consider using the species-specific models that constitute Salem: the growth models including or excluding mixture effect, the bark models, the diameter distribution models, the circumference-height relationship models, as well as the volume equations for the 12 parameterised species. Salem runs on Windows, Linux, or Mac. Its user-friendly graphical user interface makes it easy to use for non-modellers. Finally, it is distributed under a LGPL license and is therefore free and open source.

Biological insurance theory predicts that, in a variable environment, aggregate ecosystem properties will vary less in more diverse communities because declines in the performance or abundance of some species or phenotypes will be offset, at least partly, by smoother declines or increases in others. During the past two decades, ecology has accumulated strong evidence for the stabilising effect of biodiversity on ecosystem functioning. As biological insurance is reaching the stage of a mature theory, it is critical to revisit and clarify its conceptual foundations to guide future developments, applications and measurements. In this review, we first clarify the connections between the insurance and portfolio concepts that have been used in ecology and the economic concepts that inspired them. Doing so points to gaps and mismatches between ecology and economics that could be filled profitably by new theoretical developments and new management applications. Second, we discuss some fundamental issues in biological insurance theory that have remained unnoticed so far and that emerge from some of its recent applications. In particular, we draw a clear distinction between the two effects embedded in biological insurance theory, i.e. the effects of biodiversity on the mean and variability of ecosystem properties. This distinction allows explicit consideration of trade-offs between the mean and stability of ecosystem processes and services. We also review applications of biological insurance theory in ecosystem management. Finally, we provide a synthetic conceptual framework that unifies the various approaches across disciplines, and we suggest new ways in which biological insurance theory could be extended to address new issues in ecology and ecosystem management. Exciting future challenges include linking the effects of biodiversity on ecosystem functioning and stability, incorporating multiple functions and feedbacks, developing new approaches to partition biodiversity effects across scales, extending biological insurance theory to complex interaction networks, and developing new applications to biodiversity and ecosystem management.

A growing body of research suggests mixed-species stands are generally more productive than pure stands as well as less sensitive to disturbances. However, these effects of mixture depend on species assemblages and environmental conditions. Here, we present the Salem simulator, a tool that can help forest managers assess the potential benefit of shifting from pure to mixed stands from a productivity perspective. Salem predicts the dynamics of pure and mixed even-aged stands and makes it possible to simulate management operations. Its purpose is to be a decision support tool for forest managers and stakeholders as well as for policy makers. It is also designed to conduct virtual experiments and help answer research questions. In Salem, we parameterised the growth in pure stand of 12 common tree species of Europe and we assessed the effect of mixture on species growth for 24 species pairs (made up of the 12 species mentioned above). Thus, Salem makes it possible to compare the productivity of 36 different pure and mixed stands depending on environmental conditions and user-defined management strategies. Salem is essentially based on the analysis of National Forest Inventory data. A major outcome of this analysis is that we found species mixture most often increases species growth, in particular at the poorest sites. Independently from the simulator, foresters and researchers can also consider using the species-specific models that constitute Salem: the growth models including or excluding mixture effect, the bark models, the diameter distribution models, the circumference-height relationship models, as well as the volume equations for the 12 parameterised species. Salem runs on Windows, Linux, or Mac. Its user-friendly graphical user interface makes it easy to use for non-modellers. Finally, it is distributed under a LGPL license and is therefore free and open source.

The mechanistic model GO+ describes the functioning and growth of managed forests based upon biophysical and biogeochemical processes. The biophysical and biogeochemical processes included are modelled using standard formulations of radiative transfer, convective heat exchange, evapotranspiration, photosynthesis, respiration, plant phenology, growth and mortality, biomass nutrient content, and soil carbon dynamics. The forest ecosystem is modelled as three layers, namely the tree overstorey, understorey and soil. The vegetation layers include stems, branches and foliage and are partitioned dynamically between sunlit and shaded fractions. The soil carbon submodel is an adaption of the Roth-C model to simulate the impact of forest operations. The model runs at an hourly time step. It represents a forest stand covering typically 1 ha and can be straightforwardly upscaled across gridded data at regional, country or continental levels. GO+ accounts for both the immediate and long-term impacts of forest operations on energy, water and carbon exchanges within the soil–vegetation–atmosphere continuum. It includes exhaustive and versatile descriptions of management operations (soil preparation, regeneration, vegetation control, selective thinning, clear-cutting, coppicing, etc.), thus permitting the effects of a wide variety of forest management strategies to be estimated: from close to nature to intensive. This paper examines the sensitivity of the model to its main parameters and estimates how errors in parameter values are propagated into the predicted values of its main output variables.The sensitivity analysis demonstrates an interaction between the sensitivity of variables, with the climate and soil hydraulic properties being dominant under dry conditions but the leaf biochemical properties being most influential with wet soil. The sensitivity profile of the model changes from short to long timescales due to the cumulative effects of the fluxes of carbon, energy and water on the stand growth and canopy structure. Apart from a few specific cases, the model simulations are close to the values of the observations of atmospheric exchanges, tree growth, and soil carbon and water stock changes monitored over Douglas fir, European beech and pine forests of different ages. We also illustrate the capacity of the GO+ model to simulate the provision of key ecosystem services, such as the long-term storage of carbon in biomass and soil under various management and climate scenarios.

There is mounting evidence that species diversity increases the temporal stability of forest growth. This stabilising effect of diversity has mainly been attributed to species differences in their response to fluctuating environmental conditions. Interactions among individuals could also contribute to the stabilising effect of diversity by increasing the mean and reducing the variance of tree growth, however, this has never been directly demonstrated. We used tree-ring width chronologies from temperate and boreal mixed stands of Eastern Canada to identify the role of interactions among individuals in the stabilising effect of diversity on forest growth. Using neighbourhood competition index and a mixed model, we compared the effect of interspecific and intraspecific interactions on the mean and the variance of tree growth. We found that interspecific interactions are less detrimental to tree growth than intraspecific interactions. We also found that interspecific interactions buffer tree response to drought and thereby reduce the variance of tree growth. Our results indicate diversity may increase the mean and reduce the variance of tree growth through interactions among individuals. Thus, we demonstrate interactions among individuals play a role in the stabilising effect of diversity on forest growth, and in doing so, we bring to light other mechanisms of the insurance hypothesis.