Removing invasive stream macrofauna shifts non-target invertebrate mesofauna through facilitation
Data files
Oct 22, 2025 version files 122.45 KB
-
README.md
9.29 KB
-
stream.covariate_dryad.csv
4.46 KB
-
stream.invertebrate_dryad.csv
108.70 KB
Abstract
Positive interactions among non-native species can drive invasional meltdowns to the detriment of native biota. Here, we assessed whether targeted control of aquatic invasive species (AIS) can benefit native species by eliminating synergies among invaders. We did so by monitoring changes in the abundance of native and non-native benthic invertebrates following the targeted removal of non-native fish and crustaceans in 10 streams on the island of Oʻahu (Hawaiʻi, USA). Benthic invertebrate sampling was conducted in paired control reaches and removal reaches in each stream immediately following removals, one month following removals, and then at two to three-month intervals for 16 months. Temporal variation in the abundance, composition, and diversity of the native and non-native invertebrates was compared among streams and between treatment reaches using multivariate data visualization and mixed effects models. We observed both seasonal shifts in overall community composition and treatment-specific effects on the abundance of common taxa that were mediated by the number of AIS fish removed. Most notably, as removal of non-native poeciliid fish increased, we observed concordant decreases (-32 ± SE 13%) in non-native caddisfly (Cheumatopsyche analis) and increases (122 ± SE 69%) in partially native chironomid midges in the treatment reaches relative to the control reaches. Our results provide experimental evidence supporting the hypothesis that predation by introduced poeciliids on midges indirectly facilitates non-native caddisfly populations via competitive release. Our findings indicate that removal of poeciliids allows midges to outcompete non-native caddisflies and increase their abundance. Our study illustrates that targeted removal of non-native species can have cascading negative effects on other introduced taxa. Understanding trophic relationships among invasive and native species can thus enhance management efforts by maximizing benefits to native species relative to effort and cost.
Dataset DOI: 10.5061/dryad.stqjq2cgm
Description of the data and file structure
Study sites
Benthic invertebrate sampling was conducted in 10 of 12 watersheds on Oʻahu (Hawaiʻi, USA) (Fig. 2), where aquatic invasive species (AIS) removals were conducted by Moody et al. (2021). The study sites were selected to capture representative variation in hydrology, land cover, and AIS density across the island. As described in Moody et al. (2021), each removal site consisted of 350m of stream divided into five sections: a 100m upstream control, a 25m buffer, a 100m removal reach, a 25m buffer, and a 100m downstream control. Locations were chosen in part to incorporate natural dispersal obstacles (e.g., plunges, waterfalls) into the buffer zones, and the reach lengths were chosen to capture habitat variability typical of Hawaiian streams and to allow for comparison to prior studies of AIS and native stream fauna in Hawai‘i (Higashi & Nishimoto, 2007).
Watershed and in-stream characterization
Watershed characteristics, including underlying geology, land use, and annual precipitation, were quantified to determine if variation in macroinvertebrate communities corresponded to environmental differences among the study sites. Watershed characteristics were determined according to resources obtained from the Hawaiʻi Division of Aquatic Resources (DAR) and U.S. Geological Survey (USGS). Elevation maps (10 m digital resolution) derived from the National Elevation Dataset (Gesch et al., 2002) were used to obtain slope and area of each watershed. ArcGIS (ArcGIS Desktop: Release 10.1, ESRI, Redlands, CA, USA) was used to measure and quantify percentage differences in agriculture, canopy cover, development intensity (high and medium), bare land and open spaces based on classes 21-24 of the National Land Cover Dataset (2001) (Homer et al., 2007), with open spaces encompassing areas subject to light construction with vegetation planted for recreational, aesthetic, and erosion control. To encompass the net rainfall experienced in the rainy season (October – April) and dry season (May – September) (Lau & Mink, 2006), we characterized average annual precipitation using the “Online Rainfall Atlas of Hawaiʻi” (Giambelluca et al., 2013).
Benthic invertebrate sampling and characterization
Benthic invertebrates were sampled over time in 10 study sites- logistical constraints prevented sampling invertebrates in two of the 12 removal sites (Figure 2). Replicate (n = 4) benthic invertebrate samples were collected following removals using a 0.093 m2 Surber sampler with 250μm mesh in the removal reach and the upstream control reach in each of the 10 study streams. Surber samplers were chosen because the reaches were dominated by riffle and run sections featuring cobble and gravel bottom. Samples were first obtained in April 2016, within two weeks of AIS removals, again within a month of AIS removals, and subsequently in 2-5 month intervals (hereafter sampling “cycles”) from November 2016 through August 2017 (n = 255 unique sampling events). During every sampling event, each stream reach was divided into four 25 m long quadrants, with one sample taken from a randomly selected location in each quadrant.
All benthos samples were preserved in 80% ethanol and then sorted under a dissecting microscope at 40x magnification. All invertebrates were picked from the surrounding substrate and enumerated according to identifications made to the lowest possible taxonomic resolution (typically genus or species) using Merritt and Cummins (2008), Knezovic et al. (2015), Landcare Research (2017), and CPIE Project (2011). Due to the difficulties of identifying larval Hawaiian Chironomidae, all chironomid taxa were only identified to family level and were assumed to be a mix of native and introduced species, following results of previous surveys in the study region (Englund et al. 2007, Brasher et al. 2004). Notably, prior surveys of O’ahu and Maui streams found that Chironominae and Orthocladiinae dominated midge communities (O’ahu: 100%, Maui: 99.8%), and that Tanypodinae were extremely rare (O’ahu: 0%, Maui: 0.2% ), suggesting that predatory midges likely made up a marginal proportion of the samples taken in our study streams (Wolff 2012).
Files and variables
File: stream.covariate_dryad.csv
Description: This dataset contains summarized information about the geography, watershed, and environmental characteristics of the study sites. Missing values that were not able to be measured by the field teams are indicated by "n/a" where applicable.
Variables
- Latitude: Latitude of the stream reach in decimal degrees
- Longitude: Longitude of the stream reach in decimal degrees
- Site: Name of the stream
- bottom structure: Dominant benthic substrate in accordance with the Wentworth scale
- DRNAREA: Area of the site drainage basin in square kilometers
- ELEV: Elevation of the site in meters above sea level
- BSLDEM10M: Mean basin slope calculated from 10m topographic layer
- SLOP30_10M: Percentage of area with a slope of 30% or greater calculated from a 10m topographic layer
- PRECIP: Mean annual precipitation in mm
- BASINPERIM: Perimeter of the basin in km
- IMPNLCD01: % of the basin with impervious surface
- LC01DEVMD: % of the basin with medium development
- LC01DEVHI: % of the basin with high development
- LC01BARE: % of the basin with bare ground
- LC01EVERG: % of the basin with evergreen forest
- LC01CROP: % of the basin with row crop agriculture
- tempMean: average within stream water temperature
- tempCV: coefficient of variation of stream water temperature
- tempMax: maximum observed stream water temperature
- Temp10: 10th percentile of stream water temperature
- Temp50: 50th percentile of stream water temperature
- Temp90: 90th percentile of stream water temperature
- TSSave: Average total suspended solids in mg/L
- TSScv: Coefficient of variation in total suspended solids
- SRPave: Average soluble reactive phosphorus in mg/L
- SRPcv: Coefficient of variation in soluble reactive phosphorus
- NO3ave: Average nitrate concentration in mg/L
- NO3cv: Coefficient of variation in nitrate
- PreNH4Removal: Ammonium concentration in mg/L in the removal reach prior to removal
- PreNH4UpCon: Ammonium concentration in mg/L in the upstream control reach prior to removal
- PostNH4Removal: Ammonium concentration in mg/L in the removal reach after removal
- PostNH4UpCon: Ammonium concentration in mg/L in the upstream control reach after removal
- PreNO3Removal: Nitrate concentration in mg/L in the removal reach prior to removal
- PreNO3UpCon: Nitrate concentration in mg/L in the upstream control reach prior to removal
- PostNO3Removal: Nitrate concentration in mg/L in the removal reach after removal
- PostNO3UpCon: Nitrate concentration in mg/L in the upstream control reach after removal
- PreSRPRemoval: Soluble reactive phosphorus concentration in mg/L in the removal reach prior to removal
- PreSRPUpCon: Soluble reactive phosphorus concentration in mg/L in the upstream control reach prior to removal
- PostSRPRemoval: Soluble reactive phosphorus concentration in mg/L in the removal reach after removal
- PostSRPUpcon: Soluble reactive phosphorus concentration in mg/L in the upstream control reach after removal
- PreTSSRemoval: Total suspended solids concentration in mg/L in the removal reach prior to removal
- PreTSSUpCon: Total suspended solids concentration in mg/L in the upstream control reach prior to removal
- PostTSSRemoval: Total suspended solids concentration in mg/L in the removal reach after removal
- PostTSSUpcon: Total suspended solids concentration in mg/L in the upstream control reach after removal
File: stream.invertebrate_dryad.csv
Description: This dataset contains summarized information regarding the invertebrates collected and identified at each of the sampling reaches over the course of the study
Variables
- Date: The date on which sampling occurred in American format
- Season: The season when sampling occurred (Fall, Winter, Spring, Summer)
- Month: The month in which sampling occurred
- Day: The day of the month in which sampling occurred
- Year: The year in which sampling occurred
- Site: The name of the site where the sample was collected
- Treatment: The specific reach where the sample was collected, either RM (removal) or UC (upstream control)
- Replicate: The replicate number for the benthos sample taken, typically four samples were taken for each unique combination of sampling date, site, and reach.
- Sample: The full unique identifier code for a sample, combining the site name, reach code, date, and replicate number.
- OTU: The operational taxonomic unit assigned to the macroinvertebrate, e.g., the lowest taxonomic resolution the animal was identified to.
- Count: The number of individuals in the sample that were identified to that OTU. In cases of partial specimens, only heads were counted.
- Status: The geographic origin of the OTU, either I (invasive/non-native), N (native), or U (unknown).
