scholarly journals Predicting insect outbreaks using machine learning: A mountain pine beetle case study

2021 ◽  
Author(s):  
Pouria Ramazi ◽  
Mélodie Kunegel‐Lion ◽  
Russell Greiner ◽  
Mark A. Lewis
Keyword(s):  
Machine Learning ◽  
Mountain Pine Beetle ◽  
Insect Outbreaks ◽  
Mountain Pine ◽  
Pine Beetle

scholarly journals Landscape Topoedaphic Features Create Refugia from Drought and Insect Disturbance in a Lodgepole and Whitebark Pine Forest

Forests ◽  
2018 ◽  
Vol 9 (11) ◽  
pp. 715 ◽  
Author(s):  
Jennifer Cartwright
Keyword(s):  
Climate Change ◽  
Mountain Pine Beetle ◽  
Pine Forest ◽  
Boosted Regression Trees ◽  
Pine Forests ◽  
Whitebark Pine ◽  
Insect Outbreaks ◽  
Landscape Characteristics ◽  
Mountain Pine ◽  
Pine Beetle

Droughts and insect outbreaks are primary disturbance processes linking climate change to tree mortality in western North America. Refugia from these disturbances—locations where impacts are less severe relative to the surrounding landscape—may be priorities for conservation, restoration, and monitoring. In this study, hypotheses concerning physical and biological processes supporting refugia were investigated by modelling the landscape controls on disturbance refugia that were identified using remotely sensed vegetation indicators. Refugia were identified at 30-m resolution using anomalies of Landsat-derived Normalized Difference Moisture Index in lodgepole and whitebark pine forests in southern Oregon, USA, in 2001 (a single-year drought with no insect outbreak) and 2009 (during a multi-year drought and severe outbreak of mountain pine beetle). Landscape controls on refugia (topographic, soil, and forest characteristics) were modeled using boosted regression trees. Landscape characteristics better explained and predicted refugia locations in 2009, when forest impacts were greater, than in 2001. Refugia in lodgepole and whitebark pine forests were generally associated with topographically shaded slopes, convergent environments such as valleys, areas of relatively low soil bulk density, and in thinner forest stands. In whitebark pine forest, refugia were associated with riparian areas along headwater streams. Spatial patterns in evapotranspiration, snowmelt dynamics, soil water storage, and drought-tolerance and insect-resistance abilities may help create refugia from drought and mountain pine beetle. Identification of the landscape characteristics supporting refugia can help forest managers target conservation resources in an era of climate-change exacerbation of droughts and insect outbreaks.

scholarly journals Local-level emergence of network governance within the U.S. Forest Service: A case study of mountain pine beetle outbreak from Colorado, USA

2020 ◽  
Vol 118 ◽  
pp. 102204 ◽  
Author(s):  
Michelle M. Steen-Adams ◽  
Jesse B. Abrams ◽  
Heidi R. Huber-Stearns ◽  
Cassandra Moseley ◽  
Christopher Bone
Keyword(s):  
Mountain Pine Beetle ◽  
Local Level ◽  
Network Governance ◽  
Forest Service ◽  
Mountain Pine ◽  
Pine Beetle ◽  
The U.S

scholarly journals Management assessment of mountain pine beetle infestation in Cypress Hills, SK

2019 ◽  
Vol 49 (2) ◽  
pp. 154-163 ◽  
Author(s):  
Mélodie Kunegel-Lion ◽  
Rory L. McIntosh ◽  
Mark A. Lewis
Keyword(s):  
Machine Learning ◽  
Mountain Pine Beetle ◽  
Random Search ◽  
Classification Tree ◽  
Control Efficiency ◽  
Detection Strategy ◽  
Mountain Pine ◽  
Pine Beetle ◽  
Model Predictions ◽  
Current Detection

Insect epidemics such as the mountain pine beetle (MPB) outbreak have a major impact on forest dynamics. In Cypress Hills, Canada, the Forest Service Branch of the Saskatchewan Ministry of Environment aims to control as many new infested trees as possible by conducting ground-based surveys around trees infested in previous years. Given the risk posed by MPB, there is a need to evaluate how well such a control strategy performs. Therefore, the goal of this study is to assess the current detection strategy compared with competing strategies (random search and search based on model predictions via machine learning), while taking management costs into account. Our model predictions via machine learning used a generalized boosted classification tree to predict locations of new infestations from ecological and environmental variables. We then ran virtual experiments to determine control efficiency under the three detection strategies. The classification tree predicts new infested locations with great accuracy (AUC = 0.93). Using model predictions for survey locations gives the highest control efficiency for larger survey areas. Overall, the current detection strategy performs well but control could be more efficient and cost-effective by increasing the survey area, as well as adding locations given by model predictions.

Sign in / Sign up

Export Citation Format

Share Document