Aim: Disentangling how stochastic and deterministic processes contribute to variation in beta diversity is a common goal for ecologists and biogeographers. However, such studies are scarce in alpine streams, especially when different diversity facets are considered. Here, we combined different approaches to examine the drivers of taxonomic, phylogenetic and functional beta diversities and discussed how our results can inform community assembly and biodiversity conservation in Tibetan streams.

Location: Tibet Plateau

Taxon: Macroinvertebrates

Methods: We first partitioned multiple facets of beta diversity (B_total) into species replacement (B_repl) and richness difference (B_rich) as well as local (LCBD) or species (SCBD) contributions. Then, we applied ordination methods to examine the relative importance of local, climatic and spatial factors on B_total, B_repl and B_rich, respectively. We explored community assembly rules using null models based on trait and phylogeny structure.

Results: B_total displayed high values and was primarily driven by B_repl. Local, climatic and spatial factors were poor predictors of the different facets of beta diversity. Null models showed that the diversity metrics did not differ from those of null expectations, suggesting that most individual streams might be occupied by species that were merely random draws from the functional or phylogenetic pools available in this region. Partitioning beta diversity into LCBD and SCBD implied that the upper canyon streams were more unique than those at lower elevations and can be valuable for biodiversity conservation.

Main conclusions: Analyzing multiple facets of beta diversity provide important insights into community assembly that cannot be acquired by focusing on taxonomic diversity only. Using a multi-faceted approach involving species, phylogenetic and trait data, our study not only sheds light on the assembly mechanisms of macroinvertebrate communities in alpine streams but also bring inspiration for biodiversity conservation in the ‘World's Third Pole’ that is highly sensitive to global change.

In October 2015, macroinvertebrate samples were taken from 55 stream sites from the Yarlung Zangbo River Basin, of which 28 sites were located in the upper canyon and the other 27 were from the lower canyon (Fig. 1). Most of the studied streams were pristine or near pristine, with scarcely any human interference. At each site, three quantitative samples were taken from riffles with a Surber sampler (area: 0.09 m²; sieve mesh: 500 μm) and were kept in a portable fridge for subsequent processing. Benthic animals were handpicked from the sediments on a white dissecting tray and were later preserved in 70% ethanol. Based on the relevant reference books (Brinkhurst, 1986; Morse et al., 1994; Armitage et al., 2012), macroinvertebrates were identified to the lowest taxonomic level (i.e., genus or species) wherever possible in the laboratory.