Data from: Species richness of trophic guilds increases with discharge and decreases with variability in tropical river fish communities
Data files
Jul 15, 2025 version files 26.63 KB
-
README.md
5.39 KB
-
Trophic_Richnes_Flow_data_40_Dryad.xlsx
21.24 KB
Abstract
Species Area Relationships (SARs) are one of the most well-established conservation biogeography patterns, and in rivers habitat area is mediated by discharge. Species richness and river discharge have a well-established positive relationship, but how discharge affects trophic diversity is less clear. Free-flowing tropical river ecosystems are hotspots of global biodiversity, but they are under increasing threat from water resource developments which alter river discharge regimes. Here we investigate relationships between river discharge metrics and the species richness of freshwater fish trophic guilds in tropical rivers of northern Australia, using data collated from 40 catchments. We analysed relationships between the species richness of freshwater fish trophic guilds and discharge metrics including mean annual discharge (Q), mean daily dry and wet season discharge and the coefficient of variation (CVQ) of Q. Invertivores and omnivores were the most species rich trophic guilds. Our results show that the species richness of trophic guilds in north Australian freshwater fishes was correlated with multiple components of wet-dry tropical river discharge regimes. The species richness of predators, invertivores and herbivore-detritivores increased with Q and wet season discharge, whereas omnivore and invertivore richness increased with dry season discharge. Increasing variability in discharge had a negative effect on the species richness of invertivores and omnivores suggesting adverse effects of low discharge periods. We found no statistical support for the hypothesis that the slope of SARs increases with trophic level as predicted by the Trophic Island Biogeography Theory. These findings suggest that decreases in wet and dry season discharge, or increases in flow variability due to water resource development or climate change, may result in the loss of trophic diversity from tropical rivers. Our results suggest that the conservation of both wet and dry season natural flow regimes in tropical rivers will be needed to protect freshwater fish trophic diversity.
Dataset DOI: 10.5061/dryad.zkh1893nr
Description of the data and file structure
The data includes Catchments, Basins (GOC = Gulf of Carpentaria, TTS = Timor Sea, NEC = North East Coast) Trophic guild (HD = Herbivore/detritivore, OM = Omnivore, IN = Invertivore and PR = Predator) and species richness of reach trophic guild, Max discharge (MaxQ), mean annual discharge ML/d (Q), CV of daily flow (CVQ), mean daily wet season discharge (wet) and mean daily dry season discharge (dry). n_year is the number of years for records at that gauging station. Total richness for each trophic guild was calculated for each basin.
Files and variables
File: Trophic_Richnes_Flow_data_40_Dryad.xlsx
Description: Trophic_Richnes_Flow_data_40_Dryad.xlsx worksheet contains the data containing the total species richness and trophic species richness for the 4 trophic guilds by catchment.
Variables
- Catchment: total of 40
- basin: three basins
- TotalRichness: Total Species recorded in the catchment
- HD: Herbivore/detritivore richness
- OM: Omnivore richness
- IN: Invertivore richness
- PR: Predator richness
- MaxQ: Max discharge
- Q: mean annual discharge ML/d
- CVQ: CV of daily flow
- No Flow
- Wet: mean daily wet season discharge
- Dry: mean daily dry season discharge
- n_years: number of years for records at that gauging station
Code/software
We used the ‘lme4’ package in R studio (version 2023.12.1 Build 402 for plots and Generalized Linear Mixed Effects Models.
The codes and script used were: Models were fitted using the lmer() function from the lme4 package in R. Fixed effect slopes, standard errors, and p-values were extracted using broom.mixed::tidy(), and model fit was assessed using marginal R² values (performance::r2()) and Akaike Information Criterion (AIC). All trophic richness data were log transformed before modelling.
Each model was in the basic form: Richnesslog ~ Hydrologylog+ (1|basin)
Total Community Richness
- TotalRichness_log ~ Q_log + (1 | basin)
- TotalRichness_log ~ Wet_log + (1 | basin)
- TotalRichness_log ~ Dry_log + (1 | basin)
- TotalRichness_log ~ CV_log + (1 | basin)
Herbivore/Detritivore Richness
- HD_log ~ Q_log + (1 | basin)
- HD_log ~ Wet_log + (1 | basin)
- HD_log ~ Dry_log + (1 | basin)
- HD_log ~ CV_log + (1 | basin)
Omnivore Richness
- OM_log ~ Q_log + (1 | basin)
- OM_log ~ Wet_log + (1 | basin)
- OM_log ~ Dry_log + (1 | basin)
- OM_log ~ CV_log + (1 | basin)
Invertivore Richness
- IN_log ~ Q_log + (1 | basin)
- IN_log ~ Wet_log + (1 | basin)
- IN_log ~ Dry_log + (1 | basin)
- IN_log ~ CV_log + (1 | basin)
Predator Richness
- PR_log ~ Q_log + (1 | basin)
- PR_log ~ Wet_log + (1 | basin)
- PR_log ~ Dry_log + (1 | basin)
- PR_log ~ CV_log + (1 | basin)
River flow metrics were calculated for each catchment. Initial calculations included the annual mean, standard deviation, the coefficient of variation, quartiles of flow, the percent completeness of the hydrograph, and the total number of zero flow days across the entire record. These calculations were performed on the data records after the missing records were removed.
Each hydrograph was analysed by water year (starting in July) to allow the calculations to capture a single wet season (all calculations made in Python using Numpy and Pandas libraries). For each water year, multiple metrics were reported, including the maximum flow, standard deviation, total number of zero flow days, days below the 10th percentile, and days above the 10th, 30th, 50th, 70th, and 90th percentiles. The total flows and daily average flows for the wet season (November to April) and dry season (May to October) are also reported. Summaries of the catchments assessed, the number of water years contained within each dataset, and the percentage of data gaps are provided in Supplementary Tables S3-S5.
Access information
Data was derived from the following sources:
-
Surface water data were obtained from the Global Runoff Data Centre Data Portal (2023), the Bureau of Meteorology (2023), the Queensland Government (2023) and the Northern Territory government (2023) and Geoscience Australia (2004) (See Figure 1 for location of hydrographic logging stations). Where available, mean daily discharge data (Q ML/d) were used. Raw data that was available at a sub-daily timestep was converted to daily mean flow using the Pandas library in Python (McKinney, 2010). Linear interpolation was used to fill gaps of 3 days or less. Larger gaps were not filled.
-
Fish species distribution and richness data by catchment as well as mean flow data was taken from Sternberg, D., & Kennard, M. J. (2013). Environmental, spatial and phylogenetic determinants of fish life-history traits and functional composition of Australian rivers. Freshwater Biology, 58(9), 1767-1778. https://doi.org/10.1111/fwb.12166
-
Sternberg, D., & Kennard, M. J. (2014). Phylogenetic effects on functional traits and life history strategies of Australian freshwater fish. Ecography, 37(1), 54-64. https://doi.org/10.1111/j.1600-0587.2013.00362.x