British Columbia Ministry of Forests and Range/L. Maclaughlan

This web page was created as a requirement for a class project for Oregon State University’s GEO 565 GIS course. All photos included in the annotated bibliography are from the associated paper under review and unless stated are the property of the respective authors. Feel free to send any comment or suggestions to T. Hender

Advertisements
Posted in Uncategorized | Leave a comment

Annotated Bibliography

———————————————————————————————————————-

Aukema BH, Carrol AL, Zhu J, Raffa KF, Sickley TA and Taylor SW. 2006. Landscape level analysis of mountain pine beetle in British Columbia, Canada: spatiotemporal developments and spatial synchrony within the present outbreak. Ecography 29:427-441.

In this paper, Aukema et al. used aerial remote sensing surveys (12 km resolution) to assess mountain pine beetle caused tree mortality and to estimate the population density of the bark beetles. The researchers attempted to determine whether the bark beetle outbreaks originated from and epicenter and then spread out, or if the originated from multiple localized populations spatially separated from one another. Using aspatial cluster analysis, they determined that four distinct time series patterns, each with increasingly growing bark beetle populations with populations that appear to be expanding from western British Columbia in an eastward manner. Aukema et al. also examined the rate of decline in population synchrony with respect to distance among different population levels. Their result indicated that bark beetle populations are largely independent at scales greater than 200 km during non-epidemic population numbers, but may be synchronous throughout the region during large beetle outbreaks. It is hoped that data from the spatial synchrony beetle outbreaks, combined with their epicentral population pattern of expansion may help land owner and manager develop measures to help mitigate future outbreaks. In order to improve low correlation for forest parameter models representing less dominant tree species, an extirpolation technique was developed in which models are assembled for an individual species from the total parameter values of all remaining species. Overall, the new 30 m NIDRM provide a greater representation of forest stand characteristics and will provide land owners with greater information for managing against forest pathogens and insect pests.

———————————————————————————————————————-

Bentz BJ, Régnière J, Fettig CJ, Hansen EM, Hayes JL, Hicke JA, Kelsey RG, Negrón JF, Seybold SJ. 2010. Climate change and bark beetles of the Western United States and Canada: direct and indirect effects. BioScience 60(8):602-613.

Global climate change has the potential to drastically change the characteristics of forest ecosystems throughout North America. In addition, these climatic changes are likely to lead to changes in bark beetle populations and their ability to invade new territories previously made unavailable due to their temperature-dependent life-histories, which includes developmental timing and cold induced mortality. Bark beetle population dynamics are also affected by changes in host species composition. In their paper Bents et al. review the direct and indirect impacts of global climate change on various aspects of bark beetle and host tree dynamics. This information could in turn be used for mitigation effort to control or prevent future bark beetle outbreaks.

The quantitative analysis of the study was based on successful bark beetle outbreaks. Simulated climate data (1961-2100) were derived from the Canadian Regional Climate Model version 4.2.0 runs ADJ and ADL, using the IPCC A2 emissions scenarios. This resulted in projected raster dataset (45 km res) of precipitation and daily maximum and minimum temperatures, which was used to create decadal sets of 30-year normals for input into spruce and mountain pine beetle models. The bark beetle models were integrated with weather and topography models using BioSIM. Models were run 30 times for 25,000 simulation points across North America. Maps from this data were created using universal kriging. Probability values were linearized by logistic transformation before interpolation and back tronsformation. Resultant maps displayed a continuous measure of bark beetle population success for three climatic models and masked using pine habitat polygons. Using seasonality and cold-tolerance models for the mountain pine beetle, finals maps were produced showing the predicted probability of mountain pine beetle adaptive seasonality and cold survival during the three climatic periods. Results of the analysis show low predicted probability of adaptive seasonality throughout the beetle’s current range and low-temperature survival high in coastal and low elevations areas. Both models predict increased probability of bark beetle success in the current range and population expansion northward and eastward into suitable host species habitat ranges.

———————————————————————————————————————-

Chen H, Walton A. 2011. Mountain pine beetle dispersal: spatiotemporal patterns and role in the spread and expansion of the present outbreak. Ecosphere 2(6):art66. doi:10.1890/ES10-00172.1

The dispersal mechanisms and patterns of ecologically important species such as the mountain pine beetle have been difficult to predict. In an attempt to model the dispersal activity of these bark beetle, Chen and Walton develop a unique model to quantitatively estimate the short-distance dispersal (SDD) and long-distance dispersal (LDD) of the beetles at the local and regional scale. Dispersal patterns were developed using aerial surveys and measurements of the distances between a bark beetle sink patch to its nearest source patch. Using three dispersal patterns identified from the modeling, the authors determined that SDD of the beetles accounted for the majority of movement into infested and non-infested areas (96.8% and 85.3% respectively) maintaining population expansion at outbreak levels; however, they noted that LDD was a key component during the early stages of expansion into remote, non-infested regions.

———————————————————————————————————————-

Chou CY, Hedden RL, Song B, Williams TM. 2008. The simulation of southern pine beetle spot growth in Loblolly pine stands. In: Proceedings of the 6th Southern Forestry and Natural Resources GIS Conference (2008), Bettinger P, Merry K, Fei S, Drake J, Nibbelink N, and Hepinstall J, eds. Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA.

In their research, Chou et al. sought out to develop models to be used within GIS applications for monitoring the growth of bark beetle populations. Rather than relying on traditional methods of modeling bark beetle populations based on tree mortality following an outbreak, they developed a program to model southern pine beetle spot growth (based on different pine stand patterns) using ArcGIS and a regression model of beetle spot growth within loblolly pine. The resulting maps were then input into Visual Nature Studio software to create 3-D visualization of the spot growth models. Their hope is that that spot initiation and growth data may be used to model bark beetle populations according to different forest conditions. This information may then be used by land use managers to develop different silvicultural treatments in response to bark beetle outbreaks. The results of the analyses produced five different spot growth scenarios to be used in comparisons based on stand composition: mature loblolly pine stands with high, medium, or low stand density, a mixed forest stand with loblolly pine, white oak, and yellow poplar, and a natural young loblolly pine stand with high stand density.

——————————————————————————————————————–

Chou CY. 2010. Simulating and assessing the southern pine beetle spot growth on management scenarios using GIS-based model and 3-D landscape visualization. [dissertation]. [Clemson (SC)]: Clemson University. 169 pp.

Although the research in this dissertation paper involve the southern pine beetle, the use if this application may be developed for modeling mountain pine beetles in the west. In this paper, Chou lays out the foundation for a spatially explicit model he developed for simulating southern pine beetle populations and stand risk assessments. The model includes a GIS-based model program, SPBSPOT, that is used for simulating spot growth patterns of bark beetles using ArcGIS software with ArcObject and Visual Basic for Application. The SPBSPOT model is composed of four subroutines (specified stand generation, spot growth simulation, affected area simulation, and buffer strip simulation) with individual interfaces developed for each subroutine to design management prescriptions based on management needs (i.e., thinning, stand restoration, and stand species mixture). Using data inputs for pine stands in the southern pine beetles range, Chou’s results show that although different management practices are able to reduce the number of beetle infested trees, the overall impact on the affected area may not change.

Evaluating spot trend and salvage effects, Chou noted that the number of infested trees in higher density stands is typically greater than in lower density stands throughout the infestation period; however, affected areas are larger in lower density stands compared to those in higher density stands. As infestation increases in lower density stands, the size of the affected area tends to increase more rapidly. The number of infested trees can be reduced by thinning the stand, but, the size of the affected area may not be reduced because of a negative relationship between stand density and spot size. Evaluation of the simulations for stand ages revealed that  the number of infected trees, size  of outbreak area and expansion rate tend to be greater in mature pine stands than in younger one.

——————————————————————————————————————–

Chou CY, Song B Hedden RL, Williams TM, Culin JD, Post CJ. 2010. Three-dimensional landscape visualizations: new technique towards wildfire and forest bark beetle management. Forests 1:82-98.

In this article, Chou et al. advocate for the use of three-dimensional landscape visualizations as a tool for greater management possibilities for land managers and scientist, and for more effective communication of forest related issues to the public at large. By incorporating GIS software, remote sensing images, simulation models and other data into 3-D landscape visualizations, models with multi-spatial, multi-temporal and multi-expressional components may be produced to better convey information for bark beetle outbreaks and other land use issues. Use of 3-D landscape visualizations can also be a tool for planning, implementing and monitoring silvicultural treatments against bark beetle threats at the local scale. Much of the text in this paper is devoted to providing example of how various 3-D visualizations can be used to satisfy the needs of a variety of stakeholders when it come to land use management, particularly in regards to wildfire and insect outbreaks. Although little attention is given to the actual modeling techniques used in the examples, the paper does provide some information on the incorporation of 3-D modeling into GIS applications, the use of remote sensing and the addition of various simulation models such as the Forest Vegetation Simulator (FVS), the Fire Area Simulator (FARSITE), CLEMBEETLE, and the Landscape Disturbance and Succession Simulation Model (LANDIS). It is apparent that the tools discussed in this paper will help provide a better medium for analyzing and conveying information for forest related issues.

——————————————————————————————————————–

Coops NC, Johnson M, Wulder MA, White JC. 2006. Assessment of QuickBird high spatial resolution imagery to detect red attack damage due to mountain pine beetle infestation. Remote Sensing of Environment 103:67-80.

Remote sensing applications, including satellite imagery and aerial photography, are a common method for determining the presence of tree mortality due to insect outbreaks such as the mountain pine beetle. This is possible due to ability to identify the presence of red attack damage, the process of tree mortality after a bark beetle infestation in which the foliage begins to turn red and senesce. As new technologies are developed, there is often the need to assess their ability to function as a means for detecting and monitoring bark beetle infestations. In this paper, Coops et al. evaluated the success in using high spatial resolution QuickBird multi-spectral imagery to detect red attack damage cause by mountain pine beetles. After conducting ANOVA statistical analysis on the different spectral components measure by QuickBird–individual spectral bands, Normalized Difference Vegetation Index (NDVI) and Red-Green Index (RGI), it was determined that the Red-Green Index was the most successful in determining red attack damaged trees from healthy trees. Using the RGI, the authors sought out to develop a binary classification of the damaged and healthy trees. This classification system was compared to ground based assessment of beetle damaged trees and was determined to have significant test results in identifying and approximating red tree attack damage. The results suggest that satellite imagery from QuickBird and other high spatial resolution spectral imagery platforms may hold great value in the fight against mountain pine beetles and other detrimental herbivorous insect that are wreaking havoc in forested lands throughout western North America.

——————————————————————————————————————–

Coops NC, Wulder MA, White JC. 2006. Integrating remotely sensed and ancillary data sources to characterize a mountain pine beetle infestation. Remote Sensing of Environment 105:83-97.

In this paper, Coops et al. continued their research in identifying red tree attack damage from remote sensed data from multi-date Landsat imagery. In these new logistic regression models, the authors incorporated topography and aspect/solar radiation variable into their models to produce a probability surface of mountain pine beetle attack damage. With the forest stand and terrain characteristics, they attempted to locate forested regions that contained the greatest likelihood for tree mortality related to bark beetle outbreaks. Using three distinct decision tree models, the results of the models revealed that hill slopes and the site index (used by many land managers to rate forest quality) were the best indicators for pine stand susceptibility, while pine stem basal area, density and canopy closure were also important factors involved. The new models created here may be another tool used in the fight against mountain pine beetle infestations and for identifying particularly susceptible pine stands.

——————————————————————————————————————–

Dennison PE, Brunelle AR, Carter VA. 2010. Assessing canopy mortality during a mountain pine beetle outbreak using GeoEye-1 high spatial resolution satellite data. Remote Sensing of Environment 114: 2431-2435.

Traditional mapping techniques for identifying mountain pine beetle killed pine trees had relied on the identification of red crown damaged trees, which typically result within one year of a bark beetle infestation. However, most satellite imagery and aerial surveys do detect or account for gray crowned trees, which are the dead trees that have lost most of their foliage due to needle senescence–typically two or three years after an outbreak. The inability to properly account for the gray crown trees may lead in errors or uncertainties when attempting to quantify the rate of tree mortality in a region. It their paper, Dennison et al. used GeoEye-1 high spatial resolution data (0.5 m resolution) for mapping both the red and gray crown canopy that is indicative of tree mortality in lodgepole pine forests. The shadow-normalized green, red, and gray canopy area gathered from GeoEye-1 closely agreed with field-estimated green, red, and gray canopy area. When all plots were combined, the remotely sensed estimates of green, red, and gray cover using GeoEye-1 were within 1.7% of the field-estimated cover. These results show that the use of high spatial resolution data from platforms such as GeoEye-1 may be an important tool in monitoring mountain pine beetle activity and the resultant tree mortality and may help better guide management decisions for mitigating again future bark beetle outbreaks.

——————————————————————————————————————–

Dymond CC, Wulder MA, Shore TL, Nelson T, Boots B, Riel BG. 2006. Evaluation of risk Assessment of Mountain pine beetle infestations. Western Journal of Applied Forertry 21(1):5-13.

In this study, Dymond et al. evaluated an established risk assessment system that combines ground-based mountain pine beetle outbreak information with pine stand characteristics data. Realizing that tradition aerial survey data often lacks spatial precision and field surveys, although highly accurate, are expensive and rarely conducted of large areas, the authors wanted to explore new methods for bark beetle data collection. They decided to use helicopter surveys using global positioning systems (GPS) for data collection since gathering GPS points from above the forest canopy provides greater accuracy than ground based surveys which are subject to interference from overhead tree. The goals of this survey was to collect information on the short-term likelihood of bark beetle infestation based on stand characteristics and nearby active beetle infestations, thus providing land managers with up to date information to implement mitigation strategies such as thinning or sanitation (removal of infested trees). The risk rating model was based on model produced by Shore and Safranyik, with modified variable taken from Forest Inventory  Planning databases. Beetle damage data was collected as the location of centroids of infected tree clusters. The data produced from the models were used to create a risk assessment for the given year and compared to risk ratings the following year. Overall, the risk assessment system was able to accurately predict the risk in stands that were infested with bark beetles, but not all high risk stands were subsequently attacked the following year. By implement similar GPS based system analyses into aerial surveys for mountain pine beetle activity and forest health, land use managers may be provided with more complete information to help guide decisions on how to deal threats of attacks from mountain pine beetle infestations.

——————————————————————————————————————–

Ellenwood JR, Krist FJ. 2007. Building a nationwide 30-meter forest parameter dataset for forest health risk assessments. In: Proceedings of the ForestSat’07 Conference: Forests, remote sensing and GIS: methods and operational tools. 2007 Nov 5-7. Montpellier, France.

In the this paper, Ellenwood et al. advocate for new standards in the National Insect and Disease Risk Map (NIDRM), which provides annual risk assessment of tree mortality due to major insect pest and diseases. At the time of its publication, NIDRM maps were published with a resolution of 1 km. Although useful for surveying large areas and getting a general overview of forest health and large outbreaks of disease and insect occurrences, this level of detail was insufficient for the local and regional level for implementing outbreak mitigation strategies. To mitigate these insufficiencies, it was proposed to produce a new generation of NIDRM maps with greater focus on forest structure parameters and with greater spatial resolution. Additional information incorporated into these national maps included data from the National Land Cover Dataset (NLCD), the National Cooperative Soil Survey (NCSS), the University of Montana Numerical Terradynamic Simulation Group (NTSG) which produces the Daymet climate datasets. Forest stand characteristics models were developed by the USFS Forest Inventory and Analysis database for more than 175 species of trees, which were then input into each national data layer and analyzed using Cubist data mining software. The final outputs were converted to a 30-meter resolution spatial dataset using ERDAS Imagine software

——————————————————————————————————————–

Jewett JT, Lawrence RL, Marshall LA, Gessler PE, Powell SL, Savage SL. 2011. Spatiotemporal relationships between climate and whitebark pine mortality in the Greater Yellowstone Ecosystem. Forest Science 57(4)320-335.

In this paper, Jewett et al. analyzed the spatiotemporal relationships between climate, and white bark pine mortality as a result of mountain pine beetle attack. The author chose to study the Greater Yellowstone Ecosystem area, where white bark pine is considered a keystone species in high elevation regions. A time series of Landsat Enhanced Thematic Mapper Plus (ETM+) and Thematic Mapper (TM) imagery was used to monitor white bark pine mortality from 1999-2008. The images were geometrically and radiometrically processed to a the EWDI vegetation index, which is used for detecting conifer mortality over time. This data was then analyzed for impacts for topography, autocorrelation (spatial and temporal) monthly climate variation data during the same time frame. Data was gathered from 40,000 randomly selected points and used to create 24 regression tree models with subsets of predictor variables to analyze relationships between topography, climate and conifer mortality. Results of the analysis indicated a statistical link between changes in climate and beetle induced whitebark pine mortality within the Greater Yellowstone Ecosystem region, with important predictor variables of autocorrelation terms, indicating a strong host-tree depletion effect. Both drier and warmer climatic conditions favored increased whitebark pine mortality.

——————————————————————————————————————–

Wulder MA, White JC, Bentz, Ebata T. 2006. Augmenting the existing survey hierarchy for mountain pine beetle red-attack damage with satellite remotely sensed data. The Forestry Chronicle 82(2):187-202.

In this paper, the Robertson et al. discuss a range of issues dealing with the management perspective in dealing with the various forms of GIS and remote sensing data, and how to relate them to specific management needs, including aspects of inventory analysis, planning and forest modeling. The authors present examples from medium (Landsat) and high (IKONOS) spatial resolution imagery and discuss the need for reducing overhead costs associated with data collection and processing. Issues addressed include the potential abilities of particular data sources, associated limitations, the processing requirements or level associated with using these data, and the range of results that are attainable, in terms of accuracy. Special focus of methods discusses relate to management for mountain pine bark beetles.


——————————————————————————————————————–

Robertson C, Nelson TA, and Boots B. 2007. Mountain pine beetle dispersal: the spatial–temporal interaction of infestations. Forest Science 53(3):395-405.

Modeling the dispersal patterns of Mountain pine beetle continues to be a pervasive goal for many researchers and land managers. In this paper, Robertson et al. attempted to model the spatial-temporal nature of bark beetle dispersal mechanisms based on the distance and direction between red and green attacked tree clusters. Additional goals for their research included the determination of how dispersal changes during beetle outbreaks snd the effect of pine forest canopy structure on dispersal patterns. Bark beetle outbreak data was collected via helicopter surveys using GPS (heli-GPS). Additional data used in GIS analysis included Forest Inventory Polygons (FIP), Vegetation Resource Inventory (VRI), Biogeoclimatic Ecosystem Classification (BEC), and a digital elevation model (DEM). The spatial patterns of green trees and those showing red attack (beetle damaged) were investigated in two phases of analysis–for general trends and biogeoclimatic trends. For each phase, patterns of spot growth and expansion were determined. Results of their analysis showed that 60% of plots showed spot proliferation occurring, despite a local population of susceptible host trees within the plot. This is evidence that the mountain pine beetles are dispersing (30-50 m, up to >100m) past suitable hosts to new locations. The findings also suggest that this short-range dispersal often causes spot infestations to coalesce into larger outbreak regions.

 ——————————————————————————————————————–

Wulder MA, White JC, Grills D, T. Nelson, Coops NC, and Ebata T. 2009. Aerial overview survey of the mountain pine beetle epidemic in British Columbia: Communication of impacts. BC Journal of Ecosystems and Management 10(1):45-58.

——————————————————————————————————————–

Posted in Uncategorized | Leave a comment