BOX 1: Biodiversity-ecosystem function relationships (BEFs) across biodiversity facets
Ecologists have long sought to understand how changes in community composition and species loss of species alter the fate of ecosystem functions. Theoretical works and large-scale experiments using plant communities have provided the foundation of BEFs. For instance, many studies have investigated the relationships between plant species richness and primary production, demonstrating a positive and saturating relationship between richness and primary production (e.g., Tilman et al. 1996; Loreau 1998). The conclusions have then been extended to multiple ecosystem types (e.g., aquatic ecosystems), functional groups (e.g., consumers species), and ecosystem functions (e.g., secondary production, carbon storage or nutrient recycling) (Hooper et al.2005; Balvanera et al. 2006; Cardinale et al. 2012).
Biodiversity-ecosystem relationships are explained by several non-exclusive mechanisms, including complementarity, facilitation and sampling (or selection) effects. Complementarity among species allows species to use different resources, eventually releasing competitive interactions; facilitation occurs when species provide resources or modify habitat that benefit the others in the community; sampling effects (aka selection or dominance effect) leads to a positive effect of biodiversity on ecosystem functions because in diverse communities the probability to include a highly productive (competitive) species is higher. Interestingly, these mechanisms often led ecosystem functions to increase at low biodiversity level and then reach a plateau at higher biodiversity levels according to a saturating relationship. The probability to include species with similar functional roles is indeed higher when biodiversity is high, increasing functional redundancy among species. Contrastingly, in some cases, negative (or neutral) BEF relationships can arise (see Hagan et al. 2021 for further discussions). These particular examples suggest that in some communities, increasing diversity might actually induce negative competitive interactions among species. Finally, biodiversity has also been shown to stabilise ecosystem functions (over space and time) by buffering ecosystem variation against environmental fluctuation (the insurance hypothesis, Yachi & Loreau 1999). Richer communities displayed higher resilience after a perturbation than poorer communities, because of the presence of species with high recovery rate.
While BEF relationships have primarily been investigated at the interspecific level, diversity within species also determines ecosystem functions. Similar mechanisms are at play -such as complementarity, redundancy, sampling effects- acting here not among species but among individuals within species. Importantly, the effects of intraspecific diversity on ecosystem functioning can be as strong as those of species diversity (Raffardet al. 2019). Therefore, recent studies plead for the existence of intraspecific- BEFs. This corroborates some mechanistic models that did not initially distinguish between intra- and interspecific diversity in their formulation, and demonstrates that biodiversity lossper se alters ecosystem functions (Loreau 1998; Norberget al. 2001). These processes (complementarity, redundancy, sampling effect) can actually be transferred to gene functions, and hence directly applied to a BEF framework in which PCCGs would be the inclusive measure of biodiversity.