Researchers find a network approach to managing invasives can be a useful tool in large ecosystems and when data is limited.
Nichole Angell, a graduate research assistant from the University of Minnesota, evaluates a watercraft inspector completing a boat inspection.
Image Credit: John Gerritsen
When non-native plants and animals find a foothold in new territories where they don’t naturally live, they can cause severe economic and ecological damage. Known as invasive species, they’re difficult to control, and pose major challenges everywhere around the world. Now, a new study suggests that looking at invasions spreading across landscapes as networked systems—with patches or regions of habitat as nodes connected by pathways for invasions to spread—could improve management strategies in large, complex ecosystems and in cases where data is limited.
“We were surprised when our research showed that network methods performed as well as they did, and that they held up well even with limited information,” said Rebecca Epanchin-Niell, an associate professor in the Department of Agricultural and Resource Economics at UMD. “We showed that a network approach could provide really useful tools for managing invasive species.”
Epanchin-Niell and a team of collaborators from around the country published a research paper on their findings in the journal Nature Sustainability on July 14, 2022.
Preventing the spread of invasive species is challenging in part because the ecosystems involved can be very complex, and the scale of invasion can encompass entire watersheds, crossing multiple counties or state jurisdictions.
Mathematical modeling approaches which predict the most effective management strategies have been used to address many natural resource issues. But these approaches are often difficult to apply in large and complex systems. It can be difficult to even collect the massive amounts of data needed to model the distribution of an invasive species and the myriad human activities and environmental factors that affect their spread.
Even for well-studied invasions, however, the resources available for controlling the spread of a species are limited, and determining where to target control activities is a major challenge. Across an entire landscape, for example, managers must decide where to prioritize management dollars and efforts to minimize further invasion.
Researchers and resource managers have acknowledged that looking at complex invaded systems as networks could be helpful. Other fields as diverse as epidemiology and law enforcement have used network approaches like this to understand the spread of diseases and the illegal pet trade. But using a network approach to guide control of widespread invasions has never been fully tested against mathematical models that predict optimal management strategies.
Epanchin-Niell and her colleagues did just that for managing zebra mussels in more than 9,000 lakes across Minnesota. Zebra mussels are a highly destructive invasive species that is spread largely by boats traveling between lakes.
Scientists and resource managers already have a wealth of data about how many boats travel on and between the various lakes, and which lakes are infested with zebra mussels. Epanchin-Niell and her colleagues used the data to develop mathematical models that identified the optimal locations to place boat inspection and cleaning stations where managers could survey for mussels and prevent further spread.
Then, the team used the same data to analyze the lakes as a networked system, where invaded and uninvaded lakes were nodes connected by boat traffic, which can transport zebra mussels from lake to lake.
To compare the results, the team calculated the number of potentially contaminated boats that would have been found at the inspection stations recommended under each management approach. The network-guided inspections found as many infested boats as inspections based on the modeling strategy.
Epanchin-Niell and the research team then tested the network approach three more times with big chunks of information missing, such as how many boats were traveling between lakes or which lakes were infested with zebra mussels and which weren’t.
These evaluations simulated cases in which there was not enough data to even apply the traditional modeling approach. The researchers found that the network approach was not only possible, but it performed surprisingly well. Across the 58 counties in their test area, the median number of contaminated boats found was 80% of that found by the models when all the data was available.
“The network-based management guidance led to really good outcomes in most cases,” Epanchin-Niell said. “Not as good as if we’d had all the data, but the approach can provide valuable recommendations for guiding management.”
The researchers would like to test the network-based approaches with other invasive species and in other systems, and to see how they perform at achieving different types of management objectives. Applying these methods in large systems, with complex dynamics, or with limited data will likely offer new management opportunities.
“These are places scientists can’t use the optimal management [modeling] approach, and don’t really have good methods for developing management strategies,” Epanchin-Niell said. “A new tool like this would be very helpful.
---
This research was supported by Resources for the Future and the National Socio-Environmental Synthesis Center (SESYNC), the National Science Foundation (Award No. DBI-1639145), the Northern Research Station, and USDA Forest Service.
