INTRODUCTION
Global biodiversity is declining continuously in the Anthropocene (Dirzo et al., 2014), which has been widely attributed to conversion from natural landscapes to agriculture, pastures and urban settlements (Foley et al., 2005; Newbold et al., 2016). Recent findings revealed that intensive human land-uses cause species losses and filter specific sets of functional traits, with homogenizing effects on entire biotic communities (Gámez-Virués et al., 2015; Newbold et al., 2015, 2020). This has severely altered ecosystem functioning, which largely relies on biodiversity (Soliveres et al., 2016; Barnes et al., 2017). Although biodiversity decline is broadly attributed to intensive land-uses, current evidence of land-use impacts is mostly based on studies focused on isolated land-use types, such as urbanization (Monteiro-Júnior et al. 2014), agriculture (Gossner et al., 2016), or pasture (Wang et al., 2018). Consequently, most studies fail to capture the complexity of biodiversity responses to intensive cover of multiple land-uses that operate simultaneously in multiple ways.
Recent evidence shows that taxonomic richness responses to intensive human land-uses differ among trophic groups (Allan et al., 2014; Le Provost et al., 2021), suggesting that using taxonomic richness alone hinders generalizing overall biodiversity responses of entire communities to land-use changes. For improved clarity herein, we define ‘community’ as all the biota of an ecosystem, and ‘assemblage’ as any taxonomic subset of the ecosystem (e.g., fish, arthropods, macrophytes; Fauth et al. 1996). The problem of focusing on taxonomic richness can be mitigated by using trait-based approaches, and empirical evidence has shown that different biotic assemblages sharing similar traits often respond similarly to intensive land-use (Gámez-Virués et al., 2015; Newbold et al., 2020). For instance, intensive land-use impacts may be stronger for large apex consumers, as they are disproportionately vulnerable to human pressures (Estes et al., 2011; Enquist et al., 2020). Moreover, biodiversity changes at the level of one assemblage may cascade to other assemblages. For example, there is evidence that changes in plant diversity have strong bottom-up effects on the diversity of arthropods (Scherber et al., 2010). Similarly, subtle shifts in diversity of apex consumers (e.g., fish) may exert cascading effects on the diversity of primary consumers (Duffy et al., 2007; Antiqueira et al. 2018). Such intensive land-use impacts on the diversity of plants and apex consumers likely alter the structure of entire communities with negative consequences for ecosystem functioning.
Human land-uses have severely altered ecosystem functioning through direct and biodiversity-mediated indirect pathways (Barnes et al., 2017). Direct effects involve changes in environmental quality. This is particularly true in aquatic ecosystems where loss of riparian vegetation, decrease in depth, depletion in oxygen availability and increased nutrient inputs cause deterioration in water quality and compromised ecosystem functioning (Walker & Walters 2019). Indirect effects are more difficult to predict because they manifest via multiple biodiversity facets such as taxonomic and functional diversities (Le Provost et al., 2020; Moi & Teixeira de Mello, 2022). Given that both taxonomic richness and functional traits underlie the ability of ecosystems to maintain their functions (Leitão et al., 2018; Le Bagousse-Pinguet et al., 2019), a joint evaluation of these two facets of biodiversity is essential to understanding how intensive human land-use alters ecosystem functioning. Human land-use may weaken the linkages between organisms, threatening, in turn, the capacity of biodiversity to promote ecosystem functioning (Eisenhauer et al., 2019). Recent analyses revealed that the indirect effects of intensive human land-uses could be as strong as associated direct effects (Moi & Teixeira de Mello, 2022). Thus, there is an urgent need to disentangle the direct and indirect effects of human land-uses on ecosystem functioning.
Here, we tested how the intensification of multiple human land-uses types affects taxonomic and functional diversities of three stream assemblages (fish, arthropods and macrophytes). We further investigate the direct and indirect (biodiversity-mediated) effects of intensive land-use cover on ecosystem functioning (Figure S1). We used a database of 61 stream sites dispersed across an area of 507,003 km² in two Neotropical areas (Amazonia and Uruguay; Figure 1) that harbor exceptionally high levels of biodiversity (Antonelli et al. 2018). We sampled streams spanning different cover of land-use types: agriculture, pasture, urbanization, and afforestation (Figure 1b,c). Those four human land-uses have expanded globally, largely due to the rapid expansion of commercial monocultures such as soybean and cereals, as well as livestock grazing, cities, and eucalyptus, pine and oil palm plantations (Grimm et al., 2008; Gossner et al., 2016; Maloney et al., 2020). We decomposed functional diversity into three independent trait categories: (i) recruitment and life-history, (ii) resource and habitat use, and (iii) body size, all of which can mediate cross-assemblage responses to intensive land use (Gámez-Virués et al., 2015; Benejam et al., 2016). As a measure of ecosystem function, we focused on standing fish biomass, which is a commonly used metric of ecosystem functioning, underpinning ecosystem services, such as fisheries production and food supply (Benkwitt et al., 2020).
We accounted for the effects of climate (temperature and precipitation), local environmental factors (sediment heterogeneity and water quality deterioration), and stream morphology (depth) to test the hypothesis that taxonomic richness, functional diversity, and diversity of trait categories are negatively affected by intensive human land-use types. We also predicted that (i) fish, arthropod, and macrophytes assemblages would respond differently to different land-use types (Hughes et al., 2009); (ii) apex consumers (fish and arthropods) would be more strongly affected by intensive land-uses than primary producers (macrophytes); (iii) negative effects of intensive human land-uses on plant diversity would have cascading effects on arthropod and fish diversities, leading to declines in standing fish biomass; and (iv) indirect (biodiversity-mediated) effects of intensive land-uses on standing fish biomass would be as strong as the direct effects. Our findings suggest that intensive agriculture, pasture, and urbanization had consistently negative effects on taxonomic and functional diversities of fish, arthropod, and macrophyte assemblages, as well as declines in standing fish biomass, through both direct and indirect effects.