Patterns Of Taxonomic And Functional Diversity And Associated Community Assembly Mechanisms Of Macroinvertebrate Across The Mainstream Of The Xijiang River

Being the main part of the Pearl River,the third largest river in China,the Xijiang River harbours an extremely high array of macroinvertebrate diversity and endemism,one basal compartment of riverine ecosystem.However,its macroinvertebrate fauna are still under investigated so far,especially its taxonomic and functional diversity.Here,this study carried out a comprehensive survey of macroinvertebrate assemblages across the mainstem from the source（Maxiong Mountain,Qujing,Yunnan）to the estuary（Shapu Town,Zhaoqing,Guangdong）in2021.This study aimed to portray the spatial patterns of taxonomic and functional diversity,and identify the key drivers（local,regional and spatial factors）and ecological（deterministic vs stochastic）proccesses in regulating these patterns.The main achievements are as follows:（1）We conducted an analysis of the water quality and land use spatial patterns in the mainstem of the Xijiang River,utilizing 15 water quality parameters and 5 land use type indicators.The analysis employed multivariate statistical methods.The study revealed distinct longitudinal distribution patterns in land use types（forest,farmland,grassland,water,construction）.Specifically,farmland and grassland exhibited a decreasing trend from upper reaches to lower reaches,while water and construction showed an increasing trend.Significant differences were observed among different river segments only for forest,water,and grassland.Univariate analysis of variance indicated significant variations in some local environmental factors between the upper reaches,middle stream,and lower reaches sections.The upper reaches section had significantly higher values of mineralization,permanganate index,total nitrogen,electrical conductivity,and nitrate compared to the middle and lower reaches sections.Conversely,the lower reaches section exhibited significantly higher values of water temperature,river width,water depth,salinity,and silica content compared to the middle and upper reaches.（2）A total of 143 species of 117 genera were identified,belonging to 66 families in 21orders of 5 phyla.Among them,aquatic insects（85 species）were the specie-richest taxa,followed by mollusks（34 species）,annelids（15 species）,malacostracans（6 species）and others（3 species）.The upper harbored the richest（113）,followed by the lower（79）and the middle（61）.One marine-oriented taxon（Nereididae gen.et.sp.）and three exotic species（Pomacea sp.,Procambarus clarkii and Physa acuta）were detected.The regional mean density and biomass（wet weight）were 309.83±326.52 ind./m²and 46.21±75.73 g/m²,respectively.The community structure showed high spatial variations among the reaches,with higher density in the upper（275.5 ind./m²）and lower（276.3 ind./m²）and higher biomass（62.8 g/m²）in the upper.Only two species numerically dominated the fauna across the mainstream:Corbicula fluminea（10.1%）and Chironomus sp.1（5.7%）,but each reach had different list of it own ones.The functional feeding groups were mainly composed of collectors-gatherers（55 species）and collectors-filterers（19 species）.（3）Both deterministic（local and regional abiotic filterings）and stochastic（spatial structuring）determined the taxonomic composition.Canonical correspondence analysis（CCA）further detected key local(elevation,river width,water depth,salinity and COD_Mn),spatial（PCNM1,PCNM2,PCNM3,PCNM4,PCNM5,PCNM7,PCNM8,PCNM23,PCNM34,PCNM35,PCNM37 and PCNM39）and land use（water,forestland and farmland）factors in structuring the taxonomic composition variation.Variation partitioning analysis（VPA）detected that three ecological drivers（local,land use and spatial）jointly explained a considerable amount（22%）of the taxonomic composition variation.While the pure contribution of land use was negligible,both spatial（6%）and local（5%）factors were significant and equivalent,and their interaction role is higher（12%）.（4）Functional composition（community weighted mean trait value,CWM）highly differentiated longitudinally,with the remarkable distinction between the upper and lower reaches.Furthermore,among the 11 traits（34 modals）selected,9 traits（14 modals）showed highly among-reach differences,including voltinism（Volt2）,adult life span（Lif1 and Life3）,body size（Size2）,drift（Drif3）,swimming ability（Swim1）,rheophile（Flow1 and Flow2）,respiration type（Resp1,Resp2 and Resp3）,body shape（Shap1 and Shap2）and feeding habitats（Tro1）.The common individuals were those mainly characterized with multivoltine,fast seasonal develepment and small size,and also exhibited,to some extent,such traits as weak swimming ability,eurythermal,branchial respiration and body shape of not streamlined.There were no significant differences observed in the functional diversity of macroinvertebrates among different river segments.Correlation analysis revealed that the Functional Richness index exhibited a positive correlation with species richness and Shannon-Wiener diversity index.On the other hand,the Simpson diversity index showed a significant negative correlation with Functional Evenness and Functional Dispersion indices.Furthermore,the Shannon-Wiener diversity index demonstrated a significant positive correlation with Functional Evenness and Functional Dispersion indices.（5）Only deterministic（local and regional abiotic filterings）determined the functional composition studied.Variation partitioning analysis（VPA）detected those two ecological drivers（local and land use）jointly explained 14%of the functional composition variation,with the highly pure contribution of local（6%）than land use（1%）,and their interaction role of 6%.Redundancy analysis（RA）further detected key local（elevation,current velocity,p H,permanganate index）,and land use（construction and water）factors in structuring the functional composition variation.For single functional trait detected,drift composition was only controlled by local abiotic filtering（current velocity and salinity）.Instead,local and regional abiotic filterings joinly regulated the compositions of feeding habitats（elevation,temperature and water）,respiration type（current velocity,water depth,salinity,p H and grassland）and body shape（elevation,current velocity,salinity,grassland and water）.And deterministic and stochastic proccesses all played key roles in the compositions of adult life span（elevation,current velocity,water depth,salinity,permanganate index,construction,PCNM2,PCNM3 and PCNM6）and body size（elevation,permanganate index,construction,PCNM2,PCNM7 and PCNM13）.