CRL: Mt Graham Red Squirrel Monitoring Program - Project Data


Joint Fire Science Program - Graduate Research Innovation Award

Influence of burn severity and fuel reduction treatments on dispersal movements of an endangered forest obligate

John L. Koprowski and Melissa J. Merrick
Intensified forest disturbance from frequent and often severe wildfires has driven increased proactive fuel management in the form of thinning and prescribed fire in an effort to reduce fuels, maintain ecosystem functioning, and protect wildlife habitat.  While prescribed fire and fire surrogates may protect forests from catastrophic wildfire, these disturbances may alter the distribution and settlement patterns of forest obligate species, likely impacting small mammals disproportionately.  While much research is devoted to characterizing how wildfire, prescribed burn, and fuel reduction treatments impact small mammal and bird communities, less in known about how these variables influence the movement, habitat selection, and settlement patterns of forest obligate species.  Understanding how burn severity and fuel reduction treatments influence movement, habitat selection, and settlement is key as it allows us to assess how much landscape connectivity is maintained, how permeable a landscape is from the perspective of a focal species, and whether movement can be maintained among forest patches following disturbance events.   We studied natal dispersal movements, and habitat selection during exploration, dispersal, and settlement in an endangered forest obligate small mammal (the Mt. Graham red squirrel; T. h. grahamensis, hereafter MGRS) within a landscape altered by past wildfires and recent fuels reduction treatments. Our findings identify thresholds in burn severity and forest structure that promote movement and facilitate settlement, provide information on how fuel reduction treatments influence these processes, and assess the placement of current fuel reduction treatment blocks with respect to their influence on target wildlife species. To view a larger version of any graphic, click on it, then click again to close expanded image.
Final report available here: PDF

Dispersal Distance

During our study, MGRS dispersed up to 10 times farther than dispersal distances documented for red squirrels in the literature (mean = 106.8 m).  Mean straight-line dispersal distance for MGRS was 679.8 ± 1067.7 m, with juvenile males dispersing significantly farther than females (mean dispersal distance: males = 969.4 m ± 1,224.8; females = 339.0 m ± 726.4; Welch t64.8 = -2.4, p = 0.02).  All long-distance dispersers encountered 1996 and 2004 fire perimeters and fire-altered forest (Figure 6)
MGRS dispersal movements relative to burn severity.


Forest Structure

During natal exploration, dispersal, and post-settlement, MGRS consistently selected locations at higher elevation (mean = 2909 m), with greater live and total basal area (mean = 46 and 64 m2/ha, respectively) greater tree height (mean = 14 m) and variability in tree height (an indicator of older forest stands; mean = 5.81), higher canopy cover (mean = 70 %), lower slope (mean = 10 %), and lower burn severity classes (mean = < 1) compared to availability of these variables on the landscape. Examination of the distribution of forest structure and burn severity values at used locations can inform managers with regard to target thresholds supporting MGRS movement and use of forested areas (Figure 8). Mean (and median) values of forest structure and burn severity characteristic of used locations can provide baseline management criteria for maintaining landscape permeability from the perspective of MGRS, as well as promote persistence of key MGRS activity areas.
Summary of forest structure variables and burn severity classes at 10,805 used and 100,000 available locations
MGRS selection of forest structure and burn severity classes at used vs. available locations


Burn Severity

The majority of juveniles were born and settled in areas outside of the 1996 and 2004 fire perimeters.  No MGRS were born or settled in areas with burn severity classes of moderate or high burn (MTBS classes 3 & 4, respectively), whereas areas with burn severities representing none -very light burn (MTBS classes 1 & 2 respectively) appear to support MGRS settlement and reproduction.
Nearly 80% of locations MGRS used during exploration, dispersal, and post-settlement occurred in areas outside of the 1996 and 2004 fire perimeters, 19% occurred in within areas classified as MTBS no – low burn (class 1), and only 3% within areas classified as MTBS low burn (class 2).  Locations in moderate and high burn areas (MTBS classes 3 & 4, respectively) were negligible.
Burn severity in MGRS natal and settlement areas with proximity to fuel reduction treatment blocks
Summary of settlement by burn severity class and proximity to fuel reduction areas
Treatment area descriptions with burn severity class
Download entire dataset: forest structure and burn severity at used and available locations
MGRS used locations relative to burn severity


Fuel Reduction Treatments

In 2011, 3 individuals settled in areas adjacent to fuel reduction treatment blocks (Animals 993, 994, and 997), but these individuals settled and died before fuel treatment activities commenced.  In 2013, we radio-collared 15 juveniles whose natal territory centers were within or adjacent to fuel reduction treatment blocks.  MGRS born within fuel reduction treatment blocks (n = 2; Animals 1111 and 1114) spent nearly 100% of their time within the treatment block boundaries (97.2 and 99.0% of locations respectively).  Of the remaining 13 animals born adjacent to treatment blocks, an average of 2.2 % of their locations (range 0 – 14.3 %) were located within treatment blocks and 3 individuals settled adjacent to treatment blocks. Fuel reduction treatment blocks treated in 2012-2013 were designed to comprise areas with a lower probability of use by MGRS, on the periphery of high use areas. Our radio telemetry data confirmed that these areas were used seldomly by MGRS during exploration, dispersal, and post-settlement.  Areas that received fuel reduction treatments ~ 10 years ago following wildfire were used more frequently by MGRS, especially individuals born within these treatment blocks.
Summary of animals born in or adjacent to fuel reduction treatment blocks, settlement history, and fates


Landscape Connectivity

Areas of highest current (Ln Current values > 0) within MGRS habitat represent an area of approximately 998 hectares that a dispersing juvenile MGRS may view as connected and promote long distance dispersal movements.  Despite having very limited data on habitat selection by MGRS in the act of dispersing, modeling connectivity in Circuitscape allowed us to identify likely dispersal corridors that MGRS use during long-distance movements.  Modeling landscape connectivity also facilitated the identification of potential pinch points inhibiting movement of MGRS from NW to SE in the Pinaleños (Figure 9).  Fuel reduction treatment areas, with the exception of Merrill Peak (treated ~ 10 years ago) had lower connectivity values (Ln Current) indicating that these areas somewhat less important in facilitating MGRS movements, generally falling outside of areas that are likely dispersal corridors (with the highest current values).
Download ASCII grid data for MGRS Probability of Use surface and MGRS Landscape Connectivity (Ln Current) surface in NAD83, UTM Zone 12 N.
Landscape connectivity model represented by Ln Current - warmer colors represent increased current and likely important corridors for dispersal in MGRS.

Pinch points representing likely areas that inhibit dispersal movements in MGRS

Project Photos