1. INTRODUCTION
Anthropogenic disturbances accelerate the rates of species invasion and extinction, resulting in a dramatic reconfigurationn of the current biodiversity (Toussaint et al., 2014; Villéger et al., 2014; Zhang et al., 2019). These disturbances alter not only the number of species in each locality (α-diversity) (Sax & Gaines, 2003) but also the species composition among localities (β-diversity) (Villéger et al., 2011; Olden et al., 2018; Erős et al., 2020). More importantly, the combination of species introductions and extinctions can increase (i.e., homogenization) or decrease (i.e., differentiation) the similarity among communities (McKinney & Lockwood, 1999; Olden & Poff, 2004). Changes will continue to occur as new species become established and others are lost, causing the formation of “novel communities” (Hobbs et al., 2006), which will have notable ecological and evolutionary implications (Olden et al., 2004).
Taxonomic homogenization and differentiation are typically quantified, treating all species identically despite the fact that they may have different ecological functions (Rahel, 2000; Taylor, 2010; Olden et al., 2018). Functional diversity, as an important component of biodiversity, represents the variation of functional traits of all species within a community that influences composition and ecosystem function (Tilman, 2001). Given the functional complementarity and/or redundancy among species traits, functional diversity may not necessarily coincide with the patterns of taxonomic diversity under anthropogenic disturbances (Pool & Olden, 2012; Villéger et al., 2014; Campbell & Mandrak, 2020). For instance, Villéger et al. (2014) found that functional homogenization of freshwater fishes in Europe exceeded taxonomic homogenization six times, and approximately 40% of the paired communities showed taxonomic differentiation, but functional homogenization. Furthermore, a pair of assemblages that exhibit taxonomic differentiation (i.e., an increase in dissimilarity) can actually be functionally homogenized when nonnative introduced species fill similar functional roles, such as ecological redundancy of traits into new geographical locations (Olden & Rooney, 2006; Pool & Olden, 2012). In contrast, taxonomic homogenization may be accompanied by functional differentiation due to the loss of unique species with similar functional roles (i.e., functionally redundant native species) in each community (Campbell & Mandrak, 2020). Therefore, the analysis of changes in functional dissimilarity is complementary to the assessment of changes in taxonomic dissimilarity (Villéger et al., 2014). Assessing the combination of taxonomic and functional components will permit a better understanding of the potential consequences of biodiversity change on ecosystem functioning under pressure.
Freshwater fish faunas are among the most intensively threatened by anthropogenic disturbances such as nonnative fish introductions, dam construction, habitat alteration and habitat fragmentation (Dias et al., 2017; Reid et al., 2019; Couto et al., 2021). The influence of human activities on fish diversity has become a primary environmental problem, especially in dam construction (Brewitt & Colwyn, 2020; Couto et al., 2021). Dams obstruct the dispersal and migration of organisms and modify habitat conditions, both of which are directly linked to the patterns of species speciation and invasion for the most part (Johnson et al., 2008; Petesse & Petrere, 2012; Smith et al., 2017). In addition, dams can break down the geographic constraints from physical barriers by allowing the dispersal of fish into systems outside their natural range (Vitule et al., 2012; Clavero et al., 2013) and may contribute to biotic homogenization through habitat changes and homogenization (Rahel, 2007; Vitule et al., 2012; Daga et al., 2015). Low-head dams have a hydraulic height < 15m and are typically overflow or spillway structures (Poff & Hart, 2002; Brewitt & Colwyn, 2020), which are a ubiquitous human disturbance in headwater streams (Jumani et al., 2020). Meanwhile, similar to large dams, low-head dams fragment river ecosystems, constrain the upstream movement of fish, create impoundments in place of running water, and alter local habitat (i.e., water deepening, flow slowing, and substrate size decrease) (Yan et al., 2013; Fencl et al., 2015; Hitchman et al., 2018). Until now, whether low-head dams can promote biotic homogenization/differentiation with the same ecological effect as larger dams have received less attention (Bu et al., 2017; Liu et al., 2019).
Previous studies on fish homogenization/differentiation driven by low-head dams have primarily focused on changes in taxonomic similarity of assemblages, but two important knowledge gaps remain. First, whether the taxonomic and functional components of fish assemblage similarities differ in their responses to low-head dams remain poorly understood. Only accounting for the taxonomic component may mask whether a community is functionally saturated, when among communities have the same number of species but may occupy different portions of functional space (Villéger et al., 2014). Meanwhile, many circumstances such as geographic barriers and harsh environmental conditions may obstruct whether a community become functionally saturated (Głowacki & Penczak, 2013; Mateo et al., 2017). Second, limited empirical evidence has showed that the varying role of population abundance in structuring homogenization and differentiation of fish assemblages (Cassey et al., 2008; Legendre, 2014); however, few investigations have assessed the role of species abundance in defining the trends towards homogenization/differentiation due to the limitation on sampling and accessibility for abundance data (Cassey et al., 2008). Although the incidence-based approaches may play a crucial role when communities differ mostly in species compositions and are geographically far apart on global or regional scales (Legendre, 2014), the abundance-based approaches may exhibit higher sensitivity in structuring homogeneous or heterogeneous patterns when species composition differ primarily in species abundance in small spatial extents (Cassey et al., 2008; Liu et al., 2019; Dai et al., 2020). Indeed, the population size (i.e., abundance) of the same species may differ substantially across locations, and the number of individuals is not equally distributed across different species in the same locations. Particularly at small scales, the incidence-based approaches may underestimate the functions of dominant species in biotic homogenization if species abundance data are overlooked (Liu et al., 2019; Dai et al., 2020).
In the present study, based on the data collected from 53 impoundments created by low-head dams (i.e., the treatment sites) and 53 free-flowing segments blew low-head dams (i.e., the reference sites) within the first-order streams (stream order, Strahler, 1957) in the Wannan Mountains, China, we examined the effects of low-head dams on patterns of fish faunal homogenization and differentiation in headwater streams based on abundance data. We aimed to (1) quantify the spatial changes in taxonomic and functional similarity of fish assemblages driven by low-head dams and (2) examine whether the direct of changes in taxonomic and functional similarity of fish assemblages differ in their responses to low-head dams.