There are significant investments by states and resource agencies in the northeast U.S. for invasive aquatic species monitoring and management. These investments in jurisdictional waters help maintain their use for drinking, industry, and recreation. It is essential to understand the risks from invasive species, because once established, species can be costly to society and difficult or impossible to control. Identifying which species are most likely to move into a new region and cause harmful impacts can aid in preventing introductions and establishment. This is especially important in response to climate change as habitats potentially become usable to previously range-restricted species. Currently, hundreds of invasive aquatic [...]
Summary
There are significant investments by states and resource agencies in the northeast U.S. for invasive aquatic species monitoring and management. These investments in jurisdictional waters help maintain their use for drinking, industry, and recreation. It is essential to understand the risks from invasive species, because once established, species can be costly to society and difficult or impossible to control. Identifying which species are most likely to move into a new region and cause harmful impacts can aid in preventing introductions and establishment. This is especially important in response to climate change as habitats potentially become usable to previously range-restricted species.
Currently, hundreds of invasive aquatic species occur in the southeast and the western U.S. and can potentially move into the northeast region. This project will help guide future monitoring efforts and bring attention to high-risk areas that could be invaded by southern and western invasive aquatic species. The research team will select 100 invasive species based on input from a regional stakeholder workshop to ensure that priority management species are considered. Then, the team will model the spread of invasive species under future climate change scenarios to understand where they will spread and when they are expected to arrive. Early detection and rapid response are essential to minimize the impact of invasive species, and this research is a critical first step to ensure that these responses are informed and based on the best available science.
Public interface on the NAS Database to allows to interact with the created modeled risk assessments
repository
USGS NAS Database
webToolMaintenanceAndSupport
The WARC Computer Application Team will build the public interface and perform numerous checks on the quality of the web interface (including security) that meet the requirements of USGS and DOI.
languages
TBD
restrictions
None
backupAndStorage
During the project, all processing stages of this data product will be stored and backed up on the NAS Database remote server at USGS Earth Resources Observation and Science (EROS) Center.
name
Software for Assisted Habitat Modeling
dataProduct
metadata
FGDC
exclusiveUse
None.
description
Tabular summaries for each species that includes summary metrics of each area (e.g., HUC12 or units identified by managers in workshops or other communications) and summaries for each area that includes summary metrics of for each species. Columns in tables would include information such as distance to known occurrence, estimated suitable habitat at time X, and risk score.
repository
ScienceBase and served on the web tool
qualityChecks
These will be simple summaries from the other products described above, which have their own quality checks. We will review scripts to summarize data.
format
The data will csv tables in long format, one for species and one for areas.
restrictions
None. This will be public facing outreach to stakeholders.
backupAndStorage
During the project, all processing stages of this data product will be stored and backed up on local network drives at USGS FORT.
dataManagementResources
These data are summary tables of the spatial data products, and will be created at the end of the project. Data management activities for this product will include QA/QC of code to generate the summaries and xml metadata. Estimated time includes 2 weeks of Engelstad.
volumeEstimate
Estimate the volume of information generated: ~1 GB.
dataProcessing
We have a similar workflow developed for INHABIT, as referenced elsewhere, and will modify this workflow to meet the needs of this project.
name
Suitability Maps
existingInput
fees
None.
description
National Hydrography Dataset Plus is a national geospatial surface water framework. Geospatial analysts and modelers use this framework to support water resources applications.
source
Horizon Systems Corporation contracted through USGS and EPA (https://nhdplus.com/NHDPlus/)
qualityChecks
This is a published data set that has its own checks that we will be using.
citation
U.S. Geological Survey, 2017, National Hydrography Dataset Plus High Resolution (NHDPlus HR) - USGS National Map Downloadable Data Collection https://www.sciencebase.gov/catalog/item/57645ff2e4b07657d19ba8e8
format
Spatial data set.
restrictions
No restrictions.
backupAndStorage
Raw data from dataset is stored and backed up on ScienceBase. Derived or intermediate versions of this dataset (or subsets thereof) will be stored and backed up on local network drives at USGS FORT.
volumeEstimate
< 1 TB
dataProcessing
Used to define where waterbodies are on the landscape.
name
Invasive species occurrence data
history
2022-06-08 09:26:53 MDT: phase Approved DMP
model
modelVersion
SAHM v.2_2_1
description
Produce and visualize correlative species distribution models including data input, processing, selection, model fitting, and visualizing outputs using machine learning algorithms (boosted regression trees, generalized linear models, multiadaptive regression splines, Maxent, and randomForest).
source
Morisette et al. 2013. VisTrails SAHM: visualization and workflow management for species habitat modeling. Ecography 36:129-135.; https://www.sciencebase.gov/catalog/item/503fbe63e4b09851b69ab463
modelInputs
Species locations and potential predictor variables as rasters.
calibrationDetails
All models will be assessed with the follow two methods. First, a 70/30 split will divide occurrence data into training and testing splits, accordingly. Second, within the training split, a 10-fold internal cross-validation will aid model fitness and calibration using a number of evaluation metrics (e.g., AUC, AUC-PR, Boyce Index) and inspection of predictor response curves.
modelOutputs
Relative habitat suitability models applied to current and future predictor layers.