Hydromorphological (HM) assessment based on field survey (SIHM-TP) and remote sensing (SIHM-DZ) HM data was compared, and relationships between environmental variables of different spatial scales and benthic invertebrate (BI) and fish assemblages were studied in large rivers in Slovenia. BI assemblages were sampled at 83, and fish assemblages at 96 sampling sites. For each sampling site, data on variables of natural characteristics and stressors (HM characteristics, land use) were obtained through field surveys at the local scale (reach) and through remote sensing at the local (reach, segment) and regional scale (catchment). Correlation analysis (Spearman correlation) revealed good relationships between assessment of HM status of large rivers based on SIHM-TP and SIHM-DZ. Relationships between biotic assemblages and environmental variables/groups of environmental variables were analysed using canonical correspondence analysis (CCA)/partial CCA (pCCA). HM and riparian land use variables obtained at the segment scale explained a higher percentage of BI (1000 m) and fish variability (5000 m) compared to HM and riparian land use variables obtained at the reach scale (500 m). BI variability was best explained by SIHM-DZ indices based on HM data obtained at 1000 m river segments (8,9 %), while fish variability was best explained by SIHM-DZ indices based on HM data obtained at 5000 m river segments (10,4 %). Variability of BI and fish assemblages explained by local and regional stressors was well separated in the variance partitioning procedure (pCCA). Local and regional stressors explained a higher percentage of fish variability (43,5 %) compared to BI variability (19,9 %). BI variability was best explained by stressors of the catchment scale (6,8 %), while fish variability was best explained by stressor of the segment scale (15,3 %). Considering the effects of natural characteristics as covariables, local stressors explained a higher percentage of fish variability (25 %) compared to regional stressors (6,6 %), while the percentage of BI variability explained by local vs. regional stressors was comparable (4,6 vs. 3,4 %). This study emphasizes the difference in important stressors and the spatial scales on which they effect BI and fish assemblages in Slovenian large rivers.
|