Data from: Freshwater fish functional diversity shows diverse responses to human activities, but consistently declines in the tropics
Data files
May 05, 2025 version files 65.43 KB
-
README.md
5.53 KB
-
support_data_FD_meta_analysis.xlsx
59.91 KB
Abstract
Freshwater environments are intertwined with human activities and the consequence has been environmental degradation and biodiversity loss. Fish provide key ecological and economic benefits, and fish abundance and diversity can be affected by human activities resulting in functional diversity (FD) changes that might scale up to ecosystem impacts. Changes in FD can be expressed by quantifying its three main FD components temporally and/or spatially: richness, regularity, and divergence. There is no consensus about how human activities affect the main components of FD. In addition, human activities might affect the functional diversity of communities differently in temperate and tropical regions because of differences in the regional pools and the distribution of functional traits. Here, using a meta-analytical approach, we assess how different human activities (e.g., deforestation, invasion, reservoirs) in freshwater systems affect FD components in fish communities. We compiled information from 2012 to 2023, and we found highly idiosyncratic patterns globally, but consistent loss of functional richness and regularity in face of human activities in the tropics. This idiosyncrasy could be related to high environmental heterogeneity or the multiple ways in which communities can be affected by human activities, or the distribution of functional uniqueness and redundancy. The reduction of functional diversity is concerning since human activities are removing specific functions from natural environments and results in the dominance of traits related to generalist ecological strategies. Despite the general patterns of reduction, local features are determinants on how the community will answer to human activity and therefore we highlight the importance of understanding the environment and fauna at the local scale, and the mechanisms by which each activity might affect FD.
https://doi.org/10.5061/dryad.x0k6djhv4
Description of the data and file structure
This dataset contains the data used to compute the mixed-effect models. Each row is an observation of the meta-analysis.
The observations are categorized as Categoric or Correlation. Categoric ones have functional diversity mean, error and replica number for each group (control and treatment). Control is the group with sites under a less stressed environment and Treatment is the group with sites under a higher human pressure. Correlation ones have the value of the coefficient of correlaiton, error and replica number for the relationship "functional index vs impact variable". These values were used to calculate the effect size of each obersvation (yi) and its variance (vi). Based on this two latter values, we performed mixed-effect models to evaluate the general effect size of the human activity on fish functional diversity. We tested also how human activities change functional diversity for each different component of functional diversity (richness, regularity, divergence), in different locals (tropical and temperate), and how each human activity (e.g., deforestation, invasion, reservoir) affected functional diversity.
Files and variables
File: support_-_no_author_info.xlsx
Description: Contains data and traits sheets. Containing data used to compute the models of the article and traits used by each article. Missing values are indicated as "NA"
Variables - data
- Number
- Description: Id number
- Short_citation
- First Author and year of publication
- DOI
- Digital Object Identifier
- Climate
- One of the moderators used. The papers were categorized, based in their latitude, into tropical or temperate. If the mean latitude were between -23 and 23 it was considered tropical
- Index_classification
- Second moderator. The functional index used were classified into one of the three main components of functional diversity. Divergence, Regularity and Richness
- Major_impact
- Third moderator. The human activity pressioning the environment were categorized based on the study design
- yi
- Effect value (Hedges' G) of the observation based on the mean comparison or in the coefficient of correlation. Based on the values of mean, error, and n.
- vi
- Variance of yi
- Article_type
- If it was mean comparison (categoric) or a regression functional index vs variable of impact (correlation)
- Mean_c
- Mean of a functional index for the natural site category, or the less impacted category. This was considered the control
- Error_c
- Error of the control mean
- N_c
- Number of samples of the control
- Mean_t
- Mean of a functional index for the impacted site category. This was considered the treatment
- Error_t
- Error of the treatment mean
- N_t
- Number of samples of the treatment
- r
- Coefficient of correlation value of a functional index vs variable of impact relationship
- Error_r
- The error associated with the coefficient of correlation
- N_r
- Sample size of the regression
- Diversity_index
- Which index were used to calculate the response of the community
- Index_classification_b
- This classification considers a fourth component of the functional diversity. Isolating some functional index categorized as divergence into the redundancy component. It was not used
- Dec_latitude
- Mean latitude of the sample sites of the study
- Dec_longitude
- Mean longitude of the sample sites of the study
- n_traits
- Number of traits used to generate the functional space
- study_framework
- If the comparison was spatial (different place at the same time) or temporal (same place at different periods)
- years_passed
- The difference in years, for temporal studies, between the fish communities from the past and the current fish communities
Variables - traits
- Number
- Id number
- Short_citation
- First Author and year of publication
- Trait
- The name of the trait used. The trait names follow those used by the original author, which results in some trais appearing more than once under different names
- Categories
- The categories inside a qualitative trait
Code/software
The models and calculations were performed using the following packages in R environment:
- metaDigitize
- Used to extract data from articles' plots
- Joel L. Pick, Shinichi Nakagawa, Daniel W.A. Noble (2018). Reproducible, flexible and high throughput data extraction from primary literature: The metaDigitise R package BioRxiv, 1-25. DOI https://doi.org/10.1101/247775
- metafor
- Used to compute hedges' G and to perform the mixed-effect models
- Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of Statistical Software, 36(3), 1-48. https://doi.org/10.18637/jss.v036.i03
- compute.es
- Used to transform coefficient of correlation values into Hedges' g values
- AC Del Re (2013). compute.es: Compute Effect Sizes. R package version 0.2-2. URL https://cran.r-project.org/package=compute.es
Access information
Data was derived from the following sources:
- The dataset contains the DOI number for all papers from where data was extracted.
Gathering data
We did a systematic search on scopus and web of science for papers that evaluate any human activity impact on freshwater fish functional diversity and we ended with 41 articles after the exclusion of papers that were not adequate:
For the papers that compared functional diversity values from natural against impacted sites, we extracted mean, error and sample size for each study group
For papers that evaluated changes due human activities through a regression "functional index vs impact variable", we extracted the coefficient of correlation and the sample size. Also, we extracted information about the type of human activity and local of the study.
Processing data
With this data, we calculated the effect size through mixed-effects models to evaluate the effect of human activities in functional diversity. We also tested the functional component, local (tropical vs. temperate), and type of human activity as moderators of the analysis.