Geology topics

J.M. Graham

Publications and source records attributed to J.M. Graham.

3 recordsLinked to original sources

USGS2007

Analysis of Alaskan burn severity patterns using remotely sensed data

Wildland fire is the dominant large-scale disturbance mechanism in the Alaskan boreal forest, and it strongly influences forest structure and function. In this research, patterns of burn severity in the Alaskan boreal forest are characterised using 24 fires. First, the relationship between burn severity and area burned is quantified using a linear regression. Second, the spatial correlation of burn severity as a function of topography is modelled using a variogram analysis. Finally, the relationship between vegetation type and spatial patterns of burn severity is quantified using linear models where variograms account for spatial correlation. These results show that: 1) average burn severity increases with the natural logarithm of the area of the wildfire, 2) burn severity is more variable in topographically complex landscapes than in flat landscapes, and 3) there is a significant relationship between burn severity and vegetation type in flat landscapes but not in topographically complex landscapes. These results strengthen the argument that differential flammability of vegetation exists in some boreal landscapes of Alaska. Additionally, these results suggest that through feedbacks between vegetation and burn severity, the distribution of forest vegetation through time is likely more stable in flat terrain than it is in areas with more complex topography. ?? IAWF 2007.

International Journal of Wildland Fire↗

USGS2006

Landscape conditions predisposing grizzly bears to conflicts on private agricultural lands in the western USA

We used multiple logistic regression to model how different landscape conditions contributed to the probability of human-grizzly bear conflicts on private agricultural ranch lands. We used locations of livestock pastures, traditional livestock carcass disposal areas (boneyards), beehives, and wetland-riparian associated vegetation to model the locations of 178 reported human-grizzly bear conflicts along the Rocky Mountain East Front, Montana, USA during 1986-2001. We surveyed 61 livestock producers in the upper Teton watershed of north-central Montana, to collect spatial and temporal data on livestock pastures, boneyards, and beehives for the same period, accounting for changes in livestock and boneyard management and beehive location and protection, for each season. We used 2032 random points to represent the null hypothesis of random location relative to potential explanatory landscape features, and used Akaike's Information Criteria (AIC/AICC) and Hosmer-Lemeshow goodness-of-fit statistics for model selection. We used a resulting "best" model to map contours of predicted probabilities of conflict, and used this map for verification with an independent dataset of conflicts to provide additional insights regarding the nature of conflicts. The presence of riparian vegetation and distances to spring, summer, and fall sheep or cattle pastures, calving and sheep lambing areas, unmanaged boneyards, and fenced and unfenced beehives were all associated with the likelihood of human-grizzly bear conflicts. Our model suggests that collections of attractants concentrated in high quality bear habitat largely explain broad patterns of human-grizzly bear conflicts on private agricultural land in our study area. ?? 2005 Elsevier Ltd. All rights reserved.

Biological Conservation↗

USGS2005

Natural landscape features, human-related attractants, and conflict hotspots: A spatial analysis of human-grizzly bear conflicts

There is a long history of conflict in the western United States between humans and grizzly bears (Ursus arctos) involving agricultural attractants. However, little is known about the spatial dimensions of this conflict and the relative importance of different attractants. This study was undertaken to better understand the spatial and functional components of conflict between humans and grizzly bears on privately owned agricultural lands in Montana. Our investigations focused on spatial associations of rivers and creeks, livestock pastures, boneyards (livestock carcass dump sites), beehives, and grizzly bear habitat with reported human-grizzly bear conflicts during 1986-2001. We based our analysis on a survey of 61 of 64 livestock producers in our study in the Rocky Mountain East Front, Montana. With the assistance of livestock and honey producers, we mapped the locations of cattle and sheep pastures, boneyards, and beehives. We used density surface mapping to identify seasonal clusters of conflicts that we term conflict hotspots. Hotspots accounted for 75% of all conflicts and encompassed approximately 8% of the study area. We also differentiated chronic (4 or more years of conflicts) from non-chronic hotspots (fewer than 4 years of conflict). The 10 chronic hotpots accounted for 58% of all conflicts. Based on Monte Carlo simulations, we found that conflict locations were most strongly associated with rivers and creeks followed by sheep lambing areas and fall sheep pastures. Conflicts also were associated with cattle calving areas, spring cow-calf pastures, summer and fall cattle pastures, and boneyards. The Monte Carlo simulations indicated associations between conflict locations and unprotected beehives at specific analysis scales. Protected (fenced) beehives were less likely to experience conflicts than unprotected beehives. Conflicts occurred at a greater rate in riparian and wetland vegetation than would be expected. The majority of conflicts occurred in a small portion of the study area, where concentrations of attractants existed that overlapped with bear habitat. These hotspots should be the target of management and conservation efforts that focus on removing or protecting attractants using non-lethal techniques.

Ursus↗