DISCUSSION
Our results revealed that the intensive cover of human land use (i.e., agriculture, pasture, and urbanization) is negatively associated with multiple biodiversity facets of three different stream assemblages. Although our survey did not encompass all possible local environmental factors (e.g., habitat size, drainage area, riparian condition), it showed that impacts of human land use on taxonomic and functional diversities of fish, arthropod, and macrophyte were consistent even accounting for environmental, stream morphology, and climate predictors. These results are robust and represent the first spatially extensive analysis of the response of multiple biodiversity facets across entire aquatic communities to intensive human land-use in the Neotropics. Our finding is in line with previous analyses from the terrestrial realm (Newbold et al., 2015; Gossner et al., 2016), showing the negative impacts of intensive human land-use also on hyperdiverse aquatic systems. Because diversity of fish, arthropod and macrophyte assemblages responded to different human land-use types, we believe that those different land uses may act in concert to drive biodiversity patterns across multiple assemblages. Consequently, the focus on isolated land-use types likely hinders our ability to understand and manage biodiversity response to human activities across landscapes worldwide. More studies should focus on the combined impacts of multiple human stressors if we are to effectively mitigate biodiversity losses and safeguard ecosystem functioning (e.g., Benkwitt et al., 2020)
Trait diversity has been reported to decline with increasingly intensive human land-use (Newbold et al., 2020). We demonstrated significant declines in diversity of recruitment and life history, resource and habitat use, and body size of different assemblages to intensive covers of agriculture, pasture, and urbanization. Those land uses often degrade stream environmental conditions by (i) reducing habitat and resource availability; (ii) increasing over-exploitation, mainly of large individuals; and (iii) compromising water quality via excessive inputs of nutrients, pesticides, fertilizers and sewage (Allan et al., 2005; Walker & Walters, 2019; Marques et al., 2021). Here, there was a marked decline of sediment heterogeneity, stream depth, and water quality with increasingly intensive land-uses. These human-induced stressors act as filters selecting traits affecting organism resistance to disturbance, thereby reshaping biotic community composition (Williams et al., 2020). Lower habitat availability (due to sediment simplification) and stream depth both reduce availability of feeding niches (Leitão et al., 2018, Price et al., 2019), which favors generalist consumers and filters out specialized consumers including many apex predators (Walker & Walters, 2019; Cantanhêde et al., 2021). Consequently, there was a drastic reduction in the diversity of traits related to resource and habitat-use. Environmental degradation also restricts the phenology of organisms, including life histories (Morellato et al., 2016). We found that organisms with long life spans and small geographical ranges could not persist in streams with intensive agriculture, pasture, or urbanization. Consequently, the diversity of life history traits decreased. Intensive human land-use has disproportionately strong impacts on large-sized organisms because they are more vulnerable to loss of habitat and need a greater diversity of resources, which are scarcer under high land-use intensity (Newbold et al., 2015). This is in agreement with the strong declines in body size of fish and arthropods, and macrophytes, in streams experiencing intensive agriculture, pasture and urbanization catchment coverage. Our findings suggest a loss of trait diversity in human-altered ecosystems that is likely to impair ecosystem functioning (Frainer et al., 2017; Le Bagousse-Pinguet et al., 2021).
The structural equation modelling revealed that increased cover of human land-use affected ecosystem functioning through direct and biodiversity-mediated, indirect pathways. The two pathways were consistent regardless of the study area (Amazonia and Uruguay), suggesting a broad-scale decline of standing fish biomass resulting from human land-use intensification. The direct effect of ILUC on standing biomass is intuitive because fish biomass often declines in human-dominated systems as a result of fishing pressure, pollution and eutrophication (Duffy et al., 2016). Importantly, however, intensive land-use cover had a strong negative effect on taxonomic richness and functional diversity, particularly of fish and macrophytes, which ultimately resulted in net negative effects on standing fish biomass. Considering the trophic roles of macrophytes (primary producers) and fish (apex consumers), this indicates that intensive land-use can disrupt the bottom-up and top-down control of ecosystem biomass production. Bottom-up control of biomass production has been widely reported because macrophytes are basal organisms that structure habitats and enhance biomass production (Teixeira de Mello et al., 2015; Marsh et al., 2021). Fish are major apex consumers in aquatic ecosystems, and they can maintain high biomass despite relatively high human pressures (Duffy et al., 2016). Arthropods enhanced fish taxonomic and functional diversities, which indirectly increased standing fish biomass. This suggests that arthropods indirectly affected stream functioning through fish biodiversity. Arthropods are important food resources for fish, and their diversity is often linked to greater production of fish biomass (Correa & Winemiller, 2018). Although intensive human land use had negative effects on biodiversity, this did not break down positive biodiversity-ecosystem functioning relationships. This finding suggests that biodiversity can buffer ecosystem functioning against human pressures (Isbell et al., 2015). Thus, preserving high levels of biodiversity, including both taxonomic and functional components, is essential to maintain healthy ecosystem functioning in light of increasing human pressures.
By decomposing SEM results between study areas (Amazonia and Uruguay), we found stronger associations between diversity of different assemblages in Amazonia than in Uruguay. In general, the positive association between fish and arthropod diversity was stronger in Amazonia. This can be partly explained by the higher macrophyte diversity in Amazonia, which likely facilitates coexistence between fish and arthropods via increasing habitat heterogeneity (García-Girón et al., 2020; Monato et al., 2021). The taxonomic richness and functional diversity of macrophytes were 51% and 18% higher in Amazonia (29 species; FD= 2.8) than in Uruguay (14 species; FD= 2.3). Combined with the fact that the positive association between macrophyte on arthropod diversity was also stronger in Amazonia, these findings imply a strong bottom-up effect from the primary producers, favoring positive relationships between fish and arthropod assemblages diversities.
.
Assemblage functional diversity increased with assemblage taxonomic richness, indicating relatively low functional redundancy in the study ecosystems. The low functional redundancy suggests that fish, arthropod, and macrophyte are mostly composed of taxa with sets of different traits. This agrees with the low functional redundancy that is predicted by biogeographical hypotheses for the Neotropics (see, Leitão et al., 2016; Rodrigues-Filho et al., 2018). This implies that Neotropical biodiversity is particularly vulnerable to human pressures — species loss will likely be closely accompanied by declines in functional diversity. In addition, consistent positive associations between diversities of different assemblages highlight that biodiversity is closely related in these streams. Therefore, the diversity loss of any taxonomic group would result in cascading effects on the diversity of other taxonomic groups. We draw this conclusion based on the evidence that the negative effect of intensive land-use on macrophyte diversity indirectly affected fish and arthropod diversities, which likely unraveled the interactions between these consumers (Figure 6). We argue that preserving biodiversity requires mutual conservation of different facets of biodiversity across multiple Neotropical assemblages.
The conversion of natural landscapes for human use is a global problem that has transformed Earth‘s surface (Foley et al., 2005). Our dataset revealed how intensive cover of human land-use types differentially affect the taxonomic and functional diversities of three key stream assemblages. We demonstrated that taxonomic richness, functional diversity, and diversity of trait categories of fish, arthropod and macrophyte strongly declined with intensive cover of agriculture, pasture and urbanization. Our findings indicate that biodiversity conservation strategies should focus on joint management of multiple pressures at the catchment level. Biodiversity conservation will become even more challenging in the coming decades given projected increases in human population and climate chaos (United Nations 2018). We have shown that intensive human land uses reduce standing fish biomass, which occurs both directly and indirectly (mediated by taxonomic and functional diversities), by reducing the levels of biodiversity needed to maintain this function. This illustrates that intensive human land use impairs ecosystem function through multiple biodiversity facets, which suggest that biodiversity conservation alone is unlikely to suffice for sustaining ecosystem functions if underlying human pressures are not reduced.