Ecosystems experience natural disturbances and anthropogenic impacts that affect biological communities and ecological processes. When natural disturbance modifies anthropogenic impacts, current widely used bioassessment metrics can prevent accurate assessment of biological quality.

Our aim was to assess the ability of biomonitoring metrics to detect anthropogenic impacts at both perennial and intermittent sites, and in the latter including both flowing and disconnected pool aquatic phases. Specifically, aquatic macroinvertebrates from 20 rivers were sampled along gradients of natural flow intermittence (natural disturbance) and anthropogenic impacts to investigate their combined effects on widely used river biomonitoring metrics (i.e. taxonomic richness and standard biological indices) and novel functional metrics, including functional redundancy (i.e. the number of taxa contributing similarly to an ecosystem function, here a trophic function) and response diversity (i.e. how functionally similar taxa respond to natural disturbance and anthropogenic impacts).

Our results showed that natural flow intermittence can confound river bioassessment, and that a set of new functional metrics could be used as effective alternatives to standard metrics in naturally disturbed intermittent rivers.

RSTUDIO (usage notes):

0_FD_functions.R 
# R functions to estimate functional metrics and perform null models.

0_model_functions
# R functions required to run the models and perform null models.

0_quality_funct_space_fromdist.R 
# R function for computing the quality of functional multidimensional functional spaces.It is required to estimate functional metrics and perform null models.

1_metric_preparation.R 
# R script to estimate widely used river biomonitoring metrics (i.e. taxonomic richness and standard biological indices) and novel functional metrics, including functional redundancy (i.e. the number of taxa contributing similarly to an ecosystem function, here a trophic function) and response diversity (i.e. how functionally similar taxa respond to natural disturbance and anthropogenic impacts). 

2_models.R 
# R script to run the models assessing the ability of biomonitoring metrics to detect anthropogenic impacts at both perennial and intermittent sites. It was used in both all aquatic-phases dataset and flowing-phases dataset. 

3_null_models
# R script to run the null models. It was used in both all aquatic-phases dataset and flowing-phases dataset. 

############################################################################
Input files:

bloc_e
# Number of categories of each effect trait.

bloc_r
# Number of categories of each response trait.

effect_traits
"# Trophic characteristics of each genus were gathered from the Freshwater Information Platform (http://www.freshwaterecology.info) (Moog, 2002; Schmidt-Kloiber & Hering, 2015): grazers, miners, xylophagous, shredders, gatherers, active and passive filter feeders, predators, parasites and others. Same as with response traits, for genera without traits or for taxa not identified to genus, averaged data from other genera within the same family or from family-level traits were used. This was the case for 24 taxa identified to genus: Atrichopogon, Berdeniella, Chelifera, Chironomus, Dolichopeza, Eristalis, Hexatoma, Hydrellia, Limnephilus, Limnophora, Lispe, Niphargus, Orimarga, Oxycera, Ptychoptera, Satcheliella, Scatella, Selysiothemis, Sepedon, Simulium, Stratiomys, Tipula, Tonnoriella and Wiedemannia, as well as for the Chironomidae and Ceratopogonidae identified to subfamily or tribe. "

family_vs_genera
# Family names of each used genus.

final_set
"# Final set of widely used and functional metrics, and a set of the descriptors of flow intermittence and anthropogenic impacts. It is used in both all aquatic-phases dataset and flowing-phases dataset. Hydrological variables were: the hydrological regime of each site (HR; where 0 refers to perennial rivers and 1 to intermittent ones; calculated by using the TREHS software), the number of days in the disconnected pool phase since last sample was taken (DPi; calculated from temperature data loggers), and the total number of zero-flow days (i.e. disconnected pool or dry riverbed) during the 30-week study period (ZFt; calculated from temperature data loggers). Besides, to calculate the ZFt, we used the total number of days in the disconnected pool phase (total_DP) and the total number with dry riverbed (total_dry). The extent of anthropogenic impacts at each site was measured using the number of impacts in the Mediterranean reference criteria (MRC) (Sánchez-Montoya et al., 2009). "

response_traits
"# Traits related to resilience and resistance strategies to estimate functional richness (FRic, Villéger, Mason, & Mouillot 2008) and functional dispersion (FDis, Laliberté & Legendre, 2010; hereafter, RD) for the whole community and each trophic group. Categories of response traits: asexual reproduction, with resistance forms (i.e. diapause or dormancy, cocoons), aerial respiration (i.e. spiracle, hydrostatic vesicle), flier and burrower (i.e. epibenthic) or interstitial (i.e. endobenthic) locomotion and substrate relation, less than a year life cycle duration, more than one reproduction cycles per year, aerial active and aquatic passive (i.e. drift) dissemination (see Table S1 in Supplementary Information). Response traits were compiled from Tachet, Richoux, Bournaud, and Usseglio-Polatera (2010). Same as with effect traits, for genera without traits or for taxa not identified to genus, averaged data from other genera within the same family or from family-level traits were used. This was the case for 24 taxa identified to genus: Atrichopogon, Berdeniella, Chelifera, Chironomus, Dolichopeza, Eristalis, Hexatoma, Hydrellia, Limnephilus, Limnophora, Lispe, Niphargus, Orimarga, Oxycera, Ptychoptera, Satcheliella, Scatella, Selysiothemis, Sepedon, Simulium, Stratiomys, Tipula, Tonnoriella and Wiedemannia, as well as for the Chironomidae and Ceratopogonidae identified to subfamily or tribe. "