You are at WWETAC >>>

HomeThreatsProjectsWorkshopsUpdatesToolsMaps

WWETAC Projects

Project Title: Spatial modeling of invasive plant spread on roads and river networks in Alaska

Status: Completed

Principal Investigator: Matt J. Macander, ABR Inc. Fairbanks, Alaska

Collaborators: Tricia L. Wurtz, PNW Research Station, USDA Forest Service; Brian Hay, University of Alaska Fairbanks; Chris Swingley, ABR Inc.

E-mail Contact: Matt Macander, mmacander[at]abrinc.com

Summary: Most of Alaska’s invasive plants are found only along the state's limited road system, and Melilotus alba is one of the most widely distributed invasives in the state. Recently, Melilotus has been found to have moved from roadsides to the flood plains of at least three glacial rivers. In one of these cases, Melilotus has become a major component of the flood plain vegetation of the lower Stikine River in southeast Alaska, within the Stikine-LeConte Wilderness. The presence of Melilotus on the lower Stikine River points out the vulnerability of roadless public lands in Alaska to invaders dispersing via linked road-and-river networks. Because more and more noxious species are turning up in Alaska each year, and because they are also spreading along the roadsides, it is likely that other species will follow the roads-to-rivers route that Melilotus has taken. The objective of this project is to develop a simulation model of the potential spread of an invasive plant along roads and river networks in Alaska. The model will allow us to predict the rate of spread, and the number of years until an invasive plant will reach different roadless public conservation units. It will identify certain road-river interfaces and crossings as critical control points for certain public lands conservation units. This information will provide a means of prioritizing and evaluating the effectiveness of different management responses to invasive species in Alaska. The model will highlight which public lands are most vulnerable to invasion via linked road and river networks, as well as showing which lands are least vulnerable. We will be able to use the model to test hypotheses concerning climate change and changing flood regimes, for a variety of management actions and for a variety of invasive species. Initial model development will focus on a 10,000 km² study area (100 by 100 km) comprising the rivers and roads upstream of the Kanuti National Wildlife Refuge (NWR), north of Fairbanks. This area is bounded on the east by the Dalton Highway, and on the west by the western boundary of the wildlife refuge. Kanuti NWR is entirely located in National Hydrography Data set subregion 1904, and although it has no direct road access, there are 13 major and 112 minor crossings upstream of the refuge, all on the Dalton Highway. When the model is functioning properly on this relatively small test area, we’ll expand it to larger extents, with the goal of scaling the model up to the full extent of interior and south-central Alaska.

Key Issues/Problems Addressed: Spread of invasive plants in Alaska.

Alaska has relatively few invasive plants, and most of them are found only along the state's limited road system. Melilotus alba, or sweetclover, is one of the most-widely distributed invasives in the state. Melilotus has recently moved from roadsides to the floodplains of at least three glacial rivers (Conn 2008). Though many of Alaska’s most valuable public lands are located off the road system, they may be vulnerable to invasion by species that gain access to river floodplains from upstream roadside environments.

Study Objectives and Goals: We are producing a spatial simulation model to predict possible invasive plant spread scenarios for interior and south-central Alaska. The model will provide information about which river segments and conservation units would be affected by spread of invasive plants downstream from road crossings and estimates of the timeline for invasive plant spread. The effect of preventing spread by controlling invasive plants at certain crossings can be simulated.

General Description: Model inputs include GIS layers based on the road, crossings, and downstream river segments identified in the network model we previously developed (Macander and Wurtz 2007). Parameters include the starting distribution of the invasive plant, spread rates for road and river segments, and likelihoods of spread from roadside to river floodplain at each crossing. Seed dispersal, climate and disturbance regimes are also inputs, though these may be represented indirectly.

Status: The framework for the spread model was presented at the ESRI International User Conference in August 2008. The spread model will be demonstrated at the Alaska Invasive Species Conference in Anchorage on October 21, 2008. Model code is currently being finalized, to be completed by November 2008. Results include animations of invasive plant spread under different scenarios, vulnerability estimates for downstream conservation units, prioritization for control efforts, and guidance for future research on the biology of invasive plant spread.

Products:

- Model inputs (GIS layers and tabular inputs for models)
- Model code (Python)
- Brief report including: methods to generate inputs, procedure for running model, and example results

Background Citations:

Conn, JS, et al. 2008. Arctic, Antarctic, and Alpine Research. 40(2):298-308.

Macander, MJ and TL Wurtz. 2007. A Network Model to Help Land Managers Predict and Prevent Spread of Invasive Plants from Roads to River Systems in Alaska. IPY GeoNorth Conference, Yellowknife, Northwest Territories, Canada.

Project ID: FY07BK23