- Email: [email protected]
Balancing the conservation of wildlife habitat with subsistence hunting access: A geospatial-scenario planning framework
Landscape and Urban Planning 115 (2013) 10–17
Contents lists available at SciVerse ScienceDirect
Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan
Research paper
Balancing the conservation of wildlife habitat with subsistence hunting access: A geospatial-scenario planning framework Colin S. Shanley a,∗ , Gary P. Kofinas b , Sanjay Pyare c a b c
University of Alaska Fairbanks, Institute of Arctic Biology, Resilience and Adaptation Program, 902 N. Koyukuk Dr., P.O. Box 757000, Fairbanks, AK 99775, USA University of Alaska Fairbanks, Institute of Arctic Biology, Department of Humans and Environment, 902 N. Koyukuk Dr., P.O. Box 757000, Fairbanks, AK 99775, USA University of Alaska Southeast, Environmental Sciences and Geography Program, 11120 Glacier Highway, Juneau, AK 99801, USA
h i g h l i g h t s • • • •
Contemporary subsistence hunting increasingly relies on motorized access. Motorized access used for subsistence hunting may effect wildlife habitat. Habitat conservation and hunting access can be spatially integrated with GIS. Scenario planning can evaluate wildlife habitat and hunting access quantitatively.
a r t i c l e
i n f o
Article history: Received 20 June 2012 Received in revised form 25 January 2013 Accepted 11 March 2013 Available online 9 April 2013 Keywords: Access Adaptation Land-use mapping Resilience Scenario planning Transportation planning
a b s t r a c t Increased motorized access used for subsistence hunting has created a challenge for land managers trying to balance the conservation of wildlife habitat with the greater environmental impact of motorized access. We used an interdisciplinary approach to evaluate this challenge in a case study of subsistence moose (Alces alces) hunters who used off-highway vehicles (OHVs; e.g., four-wheelers) to access remote harvest areas in Yakutat, Alaska, USA, and the conservation needs to sustain moose. We developed a resiliencebased planning framework that combined methods from wildlife ecology, land-use mapping, and scenario planning. The study started at the community level by working with local hunters to evaluate their values and goals for subsistence moose hunting, and to identify thresholds of undesired change. This process served as the basis for evaluating how four road closure scenarios would effect the distribution of moose and hunters’ access to moose harvest areas. The effect of roads and OHV routes on moose distribution was quantified in a previous study with a GIS-based resource selection function model. An index of access was quantified on a digitized map of harvest areas. The results of the scenario analyses suggest that a balance in the conservation of wildlife habitat and subsistence hunting access could be found in the spatial arrangement of routes that are outside of important moose habitat, but within reach of preferred harvest areas. This spatially explicit planning framework may prove useful in northern communities experiencing an increased use of motorized access for contemporary subsistence hunting practices. © 2013 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Background Subsistence hunting communities across high latitude regions of the world are experiencing increased social, economic, and ecological changes that require rapid adaptation (Berkes & Jolly, 2001;
∗ Corresponding author. Present address: The Nature Conservancy, Alaska Field Office, 416 Harris Street, Suite 301, Juneau, AK 99801, USA. Tel.: +1 907 523 4929; fax: +1 907 586 8622. E-mail addresses: [email protected] (C.S. Shanley), gpkofi[email protected] (G.P. Kofinas), [email protected] (S. Pyare). 0169-2046/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.landurbplan.2013.03.006
Brinkman, Kofinas, Chapin, & Person, 2007; Condon, Collings, & Wenzel, 1995; Ford, Smit, & Wendel, 2006; Gordon et al., 2007; Kofinas et al., 2010). One such adaptation is the increased use of motorized access for more efficient hunting (e.g., four-wheelers and snowmachines). The increased use of motorized access for hunting has created new challenges for land management efforts that seek to balance the conservation of wildlife habitat with the need for subsistence access (Ahlstrand & Racine, 1993; Mills & Firman, 1986; Sowl & Poetter, 2004). The conservation of wildlife habitat and subsistence hunting with motorized vehicles are often considered mutually exclusive (Gordon et al., 2007; Mills & Firman, 1986; Sowl & Poetter, 2004). A challenge in interdisciplinary research on ecosystem stewardship is to develop new and creative approaches to meet both these
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
11
Fig. 1. A map of the study area, Yakutat, Alaska, USA and the network of existing vehicle routes. Yakutat is located along the Gulf of Alaska in the northernmost corner of the Tongass National Forest.
social and ecological goals (Chapin et al., 2006; Chapin, Kofinas, & Folke, 2009). For example, a resilience-based approach that uses road and trail closure scenarios could provide an effective framework to identify balanced solutions. Simultaneously considering both the social values of stakeholders and the ecological properties of a system increases the likelihood for more robust management decisions and resilience to unforeseen events (Berkes, Colding, & Folke, 2003; Chapin et al., 2009; Cordell & Bergstrom, 1999). Using a resilience-based approach, in abstract, one searches for ways to maintain valued attributes while avoiding thresholds to change in which valued attributes are lost (Walker et al., 2002). In this study, for instance, the valued attributes were identified as the conservation of wildlife habitat and access to subsistence resources. The ecological threshold was defined as the point in which wildlife habitat is effectively lost due to high levels of motorized disturbance (Shanley & Pyare, 2011). The associated social threshold is the point at which subsistence hunters do not have sufficient access to meet subsistence resources needs (Berman & Kofinas, 2004; Brinkman, Chapin, Kofinas, & Person, 2009). Scenario analysis can be used in an interdisciplinary manner to address resource management questions about future conditions, which ultimately builds resilience into the sustainability of valued ecosystem services (Ascher, 2009; Carpenter, Bennett, & Peterson, 2006; Peterson, Cumming, & Carpenter, 2003). In general, the objective is to generate an understanding of plausible (not necessarily likely) future conditions that provide the basis for discussions about the costs and benefits associated with alternative actions or policies (Carpenter et al., 2006). Scenario analysis has been undertaken to explore a wide variety of natural resource questions. For example, Peterson, Douglas Beard, et al. (2003) used a scenario planning approach on a regional scale in the Northern Highlands Lake District, Wisconsin, to explore the future consequences of urbanization and ecological vulnerability to undesired change, such as the loss of valued fishing opportunities. Focusing on the sustainability of arctic subsistence communities, Kruse et al. (2004) used scenario analysis to explore the multiple effects of climate change, oil development, and tourism on caribou hunting. Researchers in these and other projects have taken different approaches to scenario planning based on the central question, scope of the project, and availability of data.
To balance the conservation of moose habitat with off-highway vehicles (OHVs; e.g., four-wheelers), we developed spatially explicit scenarios to evaluate effects on both moose disturbance and subsistence hunting access. We used this geospatial approach for three reasons: (1) spatially explicit scenarios with GIS improves the evaluation and analysis of potential sources of disturbance on wildlife (Johnson, Nielsen, Merrill, McDonald, & Boyce, 2006); (2) many subsistence hunters have a strong spatial and visual orientation (Tobias, 2000); and (3) GIS allowed us to integrate and synthesize a variety of data types (i.e., wildlife distribution and land-use). The use of GIS technology for evaluating social and ecological data has become increasingly important for land-use planning. For example, Sawyer, Nielson, Lindzey, and McDonald (2006) used collar location data from mule deer (Odocoileus hemionus) and GIS to evaluate the trade-offs between deer habitat and natural gas development in Wyoming, U.S. Similarly, Braund (2007) used community interviews and participatory mapping with residents to create a geodatabase of traditional land-use near Tyonek and Beluga, Alaska, to evaluate potential industrial development conflict. However, new approaches are needed to combine these types of social and ecological analyses for more integrated assessments and adaptive management (Beier, Patterson, & Chapin, 2008). In this study, we illustrate how a resilience-based approach with spatially explicit scenario planning could provide an effective framework to achieve both the goals of habitat conservation and resource access. 1.2. Case study Our case study occurred on the Yakutat Ranger District of the Tongass National Forest in southeastern Alaska, USA (Fig. 1). The location and timing of subsistence hunting has traditionally been unregulated (Mills & Firman, 1986). With the advent of OHVs, subsistence hunters incorporated them into subsistence hunting strategies because OHVs allowed more ground to be covered in search of game and greater efficiency in transporting large game species from remote harvest areas (Mills & Firman, 1986). The continual use of OHVs has visibly affected the landscape with ruts in wetland areas that remain for years, even after a single event. The prevalence of such ruts became a concern for regional land
12
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Table 1 The study framework used to conduct an integrative geospatial-scenario analysis of subsistence hunting access and wildlife habitat conservation. 1. Identify values, goals, and thresholds to undesired change with respect to subsistence hunting access and wildlife habitat conservation. 2. Develop and map feasible multiuse access scenarios (e.g., road closures) with land managers using input and review from public hearings. 3. Interview resident subsistence hunters to map preferred harvest areas, elicit access constraints, and digitize results in a GIS. 4. Develop wildlife distribution models that include a variable(s) that quantifies the spatially explicit impact of access scenarios. 5. Quantitatively evaluate and rank the scenario’s effect on the extent of accessible harvest areas and the extent of high value wildlife habitat. 6. Determine if the results of any one scenario are acceptable for subsistence hunting access and wildlife habitat conservation, as well as any other criteria (e.g., ecosystem functions, cost, etc.). If not, apply lessons learned to develop and evaluate a new suite of alternative scenarios.
managers when anadromous (salmon) streams were affected and the wildlife species were potentially displaced due to disturbance (e.g., noise) (USDA Forest Service, 2007). Regulated access on the Tongass National Forest was mandated by the 1997 Tongass Land and Resources Management Plan (TLMP). Initial public meetings that introduced an Environmental Assessment revealed that residents were concerned that restricted motorized access would impede subsistence hunting practices. The subsistence activity of greatest concern with regards to access was the harvest of moose (Alces alces). Moose have historically been harvested in remote areas and these large animals (>500 kg) can be difficult to transport before the meat spoils. However, the large amount of meat forms a significant portion of a families’ diet, making as successful hunt important. Moose were also a species that land managers were concerned would be affected by unregulated access, given OHV damage had increased in areas thought to contain the region’s best moose habitat (USDA Forest Service, 2007). Our study objective was to develop and test a geospatialscenario planning framework (Table 1) with the goal of balancing the conservation of wildlife habitat and subsistence-hunting access. This study was conducted by evaluating the effect of four road-closure scenarios on both social and ecological goals of maintaining sufficient access and conserving moose habitat. Moose distribution models (Shanley & Pyare, 2011) that accounted for OHV impact on moose habitat were used to compare moose distribution among scenarios. Likewise, parameters of subsistence-hunting practices, elicited through moose hunter interviews, were extrapolated in GIS to compare the spatial extent of subsistence hunting access. 2. Methods 2.1. Social-ecological system Yakutat, Alaska, has approximately 800 residents with a mixed cash-subsistence economy. Approximately 60% of the residents are Alaska Native, primarily Tlingit. Tlingit subsistence activities were traditionally focused around marine resources with salmon as the main source of protein (Mills & Firman, 1986). During the early 1930s, moose migrated into the Yakutat area from interior Canada and were opportunistically harvested from river shores with fishing vessels to supplement marine protein sources. There was no evidence of moose in the area prior to that time. By the late 1980s, moose was the primary source of red meat with 70% of households consuming moose, either through direct harvest or sharing networks (Mills & Firman, 1986). In the 1960s, the Yakutat area was heavily vegetated with highquality moose forage (i.e., willow) and the moose population grew steadily (Mills & Firman, 1986). During this period, the moose
Fig. 2. The number of reported moose harvested in Yakutat, Alaska, USA from 1959 to 2007 (ADFG, 2008).
population size was believed to be >2000 (Smith & Franzmann, 1979) and the human population of Yakutat was approximately 300 residents with over 80% Alaska Native (U.S. Census, 1960). The road system was limited and hunting of moose was conducted primarily from fishing vessels along river outwash areas where a harvested moose could easily be transported. In the late 1960s, the road network expanded with logging activities, allowing increased hunting opportunities, new job opportunities and a concomitant increase in the human population (Mills & Firman, 1986). Competition also increased among residents of Yakutat and other communities (e.g., Juneau, Alaska), with hunters increasingly using OHVs to access remote harvest areas (Mills & Firman, 1986). The peak harvest of 380 moose occurred in 1970 (Fig. 2). Moose became a highly valued source of protein to indigenous and nonindigenous groups, and liberal harvest limits continued despite the belief of biologists that the moose population was declining (Mills & Firman, 1986). In the early 1970s, a series of severe winters with snowfalls >7 m (NOAA, 1983) resulted in high moose mortality levels. By the mid-1970s, the moose population declined to the point that hunting was closed between 1974 and 1977 (Fig. 2). Since then, the moose population has never returned to pre-1960s levels. Captured female moose in the late-1970s showed normal pregnancy rates, although they were shown to be nutritionally stressed (Smith & Franzmann, 1979). Long-term residents suggested the nutritional stress arose from the rapid change in the region’s vegetation composition, from higher quality moose-forage communities with willow (Salix spp.) and alder (Alnus sinuate) to lower quality communities like cottonwood (Populus trichocarpa) and spruce (Picea sitchensis) (Larsen, Motyka, Freymueller, Echelmeyer, & Ivins, 2005; Mills & Firman, 1986; Shephard, 1995). A similar pattern of successional change with an increase and decline in moose population was documented in an adjacent coastal area of the Copper River Delta, Alaska (Stephenson, Van Ballenberghe, Peek, & MacCracken, 2006). The moose population of Yakutat has remained relatively stable in recent history, utilizing early successional habitats created from natural disturbance events such as spring river flooding and avalanches, with approximately 800–1000 individuals estimated by aerial surveys during 2003 and 2004 (Oehlers, 2007). 2.2. Identifying values, goals, and thresholds In the spring of 2007, we held a series of informal meetings separately with land managers and resident moose hunters, with follow-up discussions for review of findings and clarification. The purpose of the meetings was to define each group’s respective
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
13
Table 2 Categorized route lengths (km) and sum for each of the four road closure scenarios in Yakutat, Alaska, USA. Routes (km)
Low OHV High OHV All-vehicles Total OHV All routes
Fig. 3. A ball-and-cup diagram illustrating the multiple stability states and threshold in which subsistence moose hunters can have sufficient – or insufficient – access to subsistence resources to meet their household needs in Yakutat, Alaska, USA.
values, management goals, and perceived thresholds to undesired changes with respect to moose conservation and harvesting access (e.g., Berkes, Mathias, Kislalioglu, & Fast, 2001). Land managers identified their goal as the management of habitat on the Yakutat landscape to support a healthy moose population. Due to the reduced population counts of moose on the portion of the region heavily used by OHVs, they suspected OHVs were effectively lowering the region’s carrying capacity for moose (USDA Forest Service, 2007). Managers were therefore primarily interested in whether or not there was a reduced probability of moose occurrence relative to OHV activity. Resident moose hunters emphasized the importance of using OHVs to retrieve large game species from remote harvest areas, since these species were critical to meeting their economic, nutritional, and cultural needs. Hunters indicated there was a maximum or threshold distance they were willing to travel by foot to transport a moose from a harvest site to an OHV route or road. This threshold distance can be illustrated with an adapted ball-and-cup diagram from the resilience literature (e.g., Walker, Holling, Carpenter, & Kinzig, 2004) that shows alternate stability states with more or less desirable outcomes (Fig. 3). Hunters also noted that this distance was dynamic: this study would be a “snapshot” in time that represents the current social and ecological conditions. For example, with new forms of hunting innovation and technology this distance would likely change. Nevertheless, moose hunters stated for the foreseeable future they were primarily interested in maintaining sufficient OHV access to important harvest areas. 2.3. Road closure scenarios Four road-closure scenarios under consideration by the Yakutat Ranger District to meet multiple objectives (e.g., wildlife values, stream restoration, and maintenance costs) were used for comparative analyses (USDA Forest Service, 2007; Table 2). Scenario 1 maintained the status quo with unlimited OHV access. All main roads (221 km) and the extensive network of hunter-created OHV routes (302 km) across inland meadows and beaches would remain available to search for moose and transport a harvested moose back to the community. Scenario 2 retained most of the main roads (182 km) and restricted many of the hunter created OHV
Scenarios I
II
III
IV
184 118 221 302 523
54 31 182 85 267
18 21 185 39 224
21 34 182 55 237
routes to a select few (85 km) routes that could be maintained and monitored to National Forest standards (Table 2). A seasonal closure would also be put on OHV routes near tern (Onychoprion aleuticus and Sterna paradisaea) nesting habitat (May to mid-August). Hunters would be allowed to use an OHV to retrieve a harvested moose from open roads and OHV routes by obtaining a permit from the Yakutat Ranger District prior to hunting, and if they could do so without causing resource damage. Resource damage was defined as “soil displacement or cutting of living vegetation to create a path” from the road or OHV route to a harvested moose (USDA Forest Service, 2007). Hunters would also have to attain a permit from the Alaska Department of Natural Resources to cross streams used by anadromous salmon. Scenario 3 was similar to Scenario 2 in terms of the total amount of main roads (185 km), however subsistence hunting access was not given priority over national transportation planning standards (Table 2). Therefore, Scenario 3 retained fewer OHV routes than any of the other scenarios (39 km). Perhaps most importantly, hunters would also not be able to use OHVs to retrieve moose from open roads or OHV routes. Finally, Scenario 4 incorporated public comments on Scenario 2 from a meeting with the Yakutat Ranger District in 2005. This scenario resulted in fewer main roads (182 km), many of which lacked utility (e.g., old logging roads), and a greater number of designated OHV routes (55 km) (Table 2). Hunters in Scenario 4 were also allowed to use OHVs to transport a harvested moose to open roads and OHV routes, if they could do so without causing resource damage and obtain a permit prior to hunting. Seasonal closure in areas of nesting terns and a permitting process for crossing of salmon streams would also apply to Scenario 4. 2.4. Evaluating subsistence access A random sample of one-third (n = 25) of the federally registered subsistence moose hunters were interviewed to document their land-use patterns in 2007. During these 45-min interviews, hunters (n = 21 of 25) mapped their preferred harvest areas and estimated (n = 25) how far they were willing to transport a harvested moose on foot to an OHV route or road. A preferred harvest area was defined as an area currently valued for hunting moose. Each hunter delineated harvest areas with dry erase markers on a transparency over a 1.5 m x 1 m high-resolution aerial photograph of the region (approximately 2 cm = 1 km). Harvest areas for each hunter were manually digitized into GIS in the form of grids of cells (50 m resolution) designated with 1 of 2 values: 1 for harvest and 0 non-harvest. These 21 grids were collectively summed to create one grid for the sampled hunter population. For example, if the harvest areas of three hunters overlapped, cells in the composite harvest surface acquired a value of three, indicating that 14% (3 hunters/21 total hunters) used the cell area for hunting. The distance that each hunter was willing to transport a harvested moose without motorized assistance to a road or OHV route was elicited during interviews. The mean (1.5 km), lower 95% CI (0.6 km), and upper 95% CI (2.4 km) transport distances were used for comparative analysis to represent the range of hunting
14
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Table 3 Overlapping harvest area (km2 ) categories and access score in the four road closure scenarios for Yakutat, Alaska, USA. The access score was derived from multiplying the km2 of access by the percent use category, and then summing across categories for each scenario and distance. The hunter access distances are 0.6 km (“Short”), 1.5 km (“Average”), and 2.4 km (“Long”). Scenarios Hunter harvest area percent use category
I
II
Distance
5% 10% 14% 19% 24% 29% 33% Access score
III
Distance
IV
Distance
Distance
Short
Average
Long
Short
Average
Long
Short
Average
Long
Short
Average
Long
272.5 206.7 76.9 37.0 10.8 3.5 0.3 54.4
423.6 325.8 144.2 72.9 17.1 3.5 0.3 90.8
495.1 379.6 196.0 96.8 22.2 4.2 0.3 112.7
164.4 68.5 16.8 10.0 4.2 0.0 0.0 19.7
274.0 151.0 41.7 17.0 6.8 0.5 0.0 38.4
346.7 226.8 66.0 29.2 10.0 1.4 0.0 55.9
162.3 56.2 12.4 4.0 2.2 0.0 0.0 16.1
259.3 126.0 33.4 8.1 4.3 0.0 0.0 31.7
319.8 197.5 50.5 12.8 6.4 0.8 0.0 45.4
171.3 72.6 18.3 10.1 4.2 0.0 0.0 20.6
294.6 161.5 50.0 17.4 6.8 0.5 0.0 41.6
383.2 243.1 80.4 31.9 10.0 1.4 0.0 61.7
zones around roads and OHV routes in each scenario. These zones represented portions of the landscape that were considered accessible by the of resident-hunter population. An access score was then derived for each scenario and transport distance by overlaying the scenario-specific access zones on the cumulative harvest grid. First, harvest grid-cell values occurring within access zones for each road closure scenario were summed for each percent-use category. Then, we assumed a linear relationship between the relative importance of cells and the percent of hunters using the hunting area on the cumulative harvest grid. This assumption was made to weight the relative importance of harvest areas to the subsistence hunting community as a whole for scenario analysis. To calculate each scenario’s final access score, the total cell area of each percent-use category was multiplied by the percentage of hunters that used the area those cells represented, and summed across all percent-use categories.
2.5. Evaluating moose habitat A comparison of the cumulative impact of road-closure scenarios on the amount of high-probability female moose habitat (Shanley & Pyare, 2011) was developed using a cumulative habitat score for each scenario. To do this, we used GIS to map the proposed configuration of roads and OHV routes for each of the four scenarios. Each route was categorized as either Low OHV, High OHV, and All-Vehicles by size and surface wear, and then weighted with the mean number of seasonal one-way trips for each route category (Shanley & Pyare, 2011); a parameter estimated from hunter interviews. A female moose distribution model (Shanley & Pyare, 2011) was used to evaluate the effect of the four road closure scenarios on moose distribution. A female distribution model was selected due to their reproductive importance and higher likelihood of disturbance (Shanley & Pyare, 2011). The model was a spatially explicit resource selection function (RSF Design II; Johnson et al., 2006; Manly, McDonald, Thomas, McDonald, & Erickson, 2002) that incorporated a parameter for OHV activity derived from interviews, as well as habitat composition and configuration. The resulting RSF for each scenario was a mapped surface of 20 m × 20 m grid cells representing the probability of moose occurrence across the study area. We utilized an average predicted probability/suitability approach (Liu, Berry, Dawson, & Pearson, 2005) where the mean probability of occurrence was used to determine a 0.4 threshold probability value for delineating species occurrence across the study area. Cells above the threshold probability value were classified into 0.1 probability intervals (e.g., 0.4–0.5, 0.5–0.6, etc.) for scenario analysis. We assumed a linear relationship between the probability of moose
Table 4 Moose habitat (km2 ) by probability interval and habitat score in the four road closure scenarios for Yakutat, Alaska, USA. The habitat score was derived by multiplying the km2 of habitat by the probability of use interval mid-point, and then summing across probability-corrected intervals for each scenario. Moose habitat probability intervals
0.4–0.5 0.5–0.6 0.6–0.7 0.7–0.8 0.8–0.9 0.9–1.0 Habitat score
Scenarios
I
II
III
IV
231.4 196.0 72.9 23.3 7.9 0.2 283.7
263.5 218.6 75.4 23.4 7.9 0.2 312.3
263.9 222.4 75.7 23.4 7.9 0.2 314.7
262.8 218.6 75.4 23.4 7.9 0.2 312.0
occurrence and the relative importance of habitat, and the total area of habitat in each probability interval was multiplied by the average probability of the interval. This assumption was made to weight the relative importance of habitats for scenario analysis. The final habitat score was calculated from the sum of the probability-corrected habitat intervals for each road closure scenario. 3. Scenario results Scenario 1 maintained the status quo and included the greatest number of roads and OHV routes; and consistently resulted in the best scores for subsistence hunting access across the range of short, average, and long-distance hunter categories (Table 3 and Fig. 4). It also resulted in the lowest score for female moose habitat (Table 4 and Fig. 5), primarily seen in the 0.4–0.6 probability categories, because female moose were influenced by this larger road network. Scenario 2 (the U.S. Forest Service-preferred alternative) resulted in the second lowest scores for subsistence hunting access (Table 3) and a habitat score for female moose similar to Scenario 4 (Table 4). The closing of many roads and OHV routes would approximately halve the access scores of Scenario 1, but result in greater habitat area for female moose. Scenario 3, which met minimum policy requirements, resulted in the lowest score for subsistence hunting access (Table 3) and the greatest habitat score for female moose (Table 4). The low access scores and the high habitat score were the result of fewer roads and OHV routes. Scenario 4 was based on public comments on Scenario 2 and yielded the second highest scores for subsistence hunting access (Table 3) and a habitat score for female moose similar to Scenario 2 (Table 4). Scenario 4 had fewer roads but more OHV routes, which could explain why it provided for greater hunter access (Fig. 4).
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
15
Fig. 4. Map of subsistence moose hunters’ preferred harvest areas by percent use categories in Yakutat, Alaska, USA and the average distance (hashed areas) hunters are willing to transport a harvested moose from the existing network of vehicle routes in Scenario 1. Increasingly darker harvest areas represent increased hunter use. An inset (a) shows an area of extensive OHV routes in Scenario 1 and (b) shows the same visual extent and access to preferred harvest areas with fewer OHV routes in Scenario 4.
4. Discussion The results of the road closure scenarios illustrate a socialecological dynamic created by the spatial configuration of routes on moose habitat and subsistence hunting access. Comparison of the results of Scenario 1 with Scenario 4 illustrates how public comments incorporated in Scenario 4 consistently improved hunting access for short, average, and long-distance hunter categories (Table 3) with minimal impact on moose habitat (Table 4). The existing network in Scenario 1 yielded a greater degree of access but also had a noticeable disturbance effect on female moose (Table 4
and Fig. 5). Scenario 4 had fewer total routes but more harvest areas within reach of roads and OHV routes than any of the other scenarios. Maintaining the status quo in Scenario 1 without enacting new regulations would also allow future OHV route expansion that could pose a future impact to female moose habitat. The proposed regulations for Scenario 4 would ban the creation of new OHV routes but allow the retrieval of a harvested moose from designated routes, therefore limiting future impact on female moose habitat and allowing for more efficient transportation of a harvested moose. Additionally, the closure of main roads (All-Vehicle routes) seen in Scenario 1 illustrates how, kilometer-for-kilometer, the
Fig. 5. Resource selection function model of female moose distribution with the existing network of vehicle routes in Scenario 1; illustrating the impact of roads and OHV routes on female moose habitat selection in Yakutat, Alaska, USA. Increasingly darker areas represent a higher probability of moose occurrence. An inset (a) shows an area of extensive OHV routes in Scenario 1 and (b) shows the same visual extent with fewer OHV routes in Scenario 4.
16
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
closure of main roads would have resulted in greater habitat gains for moose than the closure of OHV routes (Fig. 5) due to the higher disturbance effect caused by main roads (Shanley & Pyare, 2011). Scenario 4 was ultimately selected by the Yakutat Ranger District as the adopted plan (USDA Forest Service, 2009a, 2009b). Although Scenario 4 showed promise through the incorporation of public comments, no clear “winner” emerged as the best scenario for the conservation of moose habitat and subsistence hunting access. Indeed, it may be too simplistic to utilize a linear relationship to score the extent of moose habitat and hunting access as proxies for evaluating planning scenarios. Rather, a balance in the conservation of wildlife habitat and subsistence hunting access may be achieved via the precise spatial configuration of transportation networks. More specifically, a balance could be achieved by the strategic closure of roads and OHV routes in specific areas with high probability moose habitat and the maintenance of routes that are within reach of important harvest areas. The methodology demonstrated here – the spatial integration of ecological and subsistence access – could likewise be applied to finer-scale scenario planning involving individual routes and specific harvest areas. Follow-up interviews with the Yakutat Ranger District in 2012 suggest the implementation of Scenario 4 has had successes and challenges that will likely require future amendments and finerscale analysis. While Scenario 4 has allowed for improved habitat protection in closed areas, the provision to allow OHV use for moose harvest retrieval has presented a challenge for assessing fine-scale environmental impacts adjacent to open roads and OHV routes. New studies are underway to quantify the environmental impacts using a system of hunter reporting and harvest site surveys. The land managers suggested the results of these studies will likely result in additional OHV route closings and in turn new trail openings to achieve a balance in the conservation of wildlife habitat and subsistence hunting access. The results of the scenario analyses to date should be treated within the context of the assumptions made in our spatial analyses. The distance hunters were willing to transport a harvested moose to an OHV route or road was clearly identified as an important management consideration, but simplifying it to a range of three distances for our analyses should not overshadow the complexity of this metric that depends on yearly hunting conditions (e.g., weather), and will likely change as hunters adapt to new social and ecological conditions. Similarly, the disturbance of female moose will depend on the spatial configuration of habitats (Shanley & Pyare, 2011) that will also likely change with continued successional processes (Larsen et al., 2005; Shephard, 1995; Stephenson et al., 2006). Nevertheless, the resilience-based approach using a geospatial-scenario framework provided a quantitative and interdisciplinary process for assessing future management iterations. Indeed, increased social and ecological change, e.g., climate change, will necessitate flexibility as a crucial management consideration in the development of regulations that manage subsistence hunting access. A review of other land-use studies suggests a geospatialscenario framework may prove useful in many other regions where motorized subsistence access is increasing and the conservation of wildlife habitat is a concern. Some of the first subsistence landuse studies occurred in the Canadian Arctic, documenting Native land claims in the face of industrial development (Freeman, 1976). The results of these studies demonstrated how subsistence mapping could elicit detailed use of the landscape and the distance hunters traveled to reach harvest areas. Subsequent studies have also produced species-specific harvest maps, demonstrating highvalue harvest areas that have stayed in the same locations for generations or alternatively, where social and ecological conditions have changed and required adaptation through, for instance, motorized vehicles to hunt more efficiently (Berkes et al., 1995;
Natcher, 2004; Pedersen & Coffing, 1984). The variety of subsistence land-use patterns depends on the region and species of interest, and underscores the importance of time and place specific analyses. For example, Berman and Kofinas (2004) showed that the subsistence caribou hunters of Old Crow, Yukon preferred to harvest caribou on time-tested accessible migration routes in order to save resources (i.e., time and money), when possible. When the migrations changed, additional resources had to be used on extended trips with motorboats and snowmachines to secure a sufficient harvest. A similar pattern of harvest in local areas mixed with long-distance motorized travel was shown by Natcher (2004) in the Yukon Flat communities of interior Alaska. Moose hunters of Birch Creek preferred to hunt local wetland areas, but changing interior fire regimes (natural and prescribed) caused a time lag in moose-forage availability between burns in these wetland areas. The changing distribution of moose required an additional investment of time and money to search for moose with motorboats or snowmachines. In addition to the variety of land-use patterns described in subsistence studies, it is clear that accessing and securing subsistence resources is changing, and motorized access is shaping how hunters perceive and use the landscape (Brinkman et al., 2009).
5. Conclusion The use of motorized access for the harvest of subsistence resources is likely to increase in Yakutat (USDA Forest Service, 2007), as it is in many northern subsistence communities (Berkes et al., 1995; Condon et al., 1995; Ford et al., 2006; Gordon et al., 2007). Increased integration of subsistence harvesting practices with the cash economy is likely to mean contemporary subsistence hunters will have less time to spend on the land than past generations, requiring more efficient means of accessing hunting areas. In addition, the expense of modern equipment such as snowmachines, OHVs, rifles, ammunition, and fuel, will result in an even greater involvement in the cash economy (Berman & Kofinas, 2004; Fast & Berkes, 1998). Furthermore, shifts in regional climate regimes will compound social-ecological changes due to greater investment in locating and securing the shifting distribution of resources (Chapin et al., 2004). In some regions or with some species, the influence of motorized access on wildlife habitat may be negligible (i.e., highly dispersed wildlife) or ecological changes will improve hunting opportunities. However, as this study suggests, access management plans in some regions need to be updated to better reflect changing social and ecological conditions. A resilience-based approach that incorporates geospatial-scenarios with integrative quantitative and qualitative components may provide an effective framework to balance the conservation of wildlife habitat with contemporary subsistence hunting practices.
Acknowledgements We would like to thank the hunters who participated in this study for sharing their time, knowledge, and hunting experiences. Members of the Yakutat Tlingit Tribe J. Ramos and E. Henninger, and B. Lucey of the Yakutat Salmon Board provided us with immense local insight. A.L. Jacob made helpful comments on the developing manuscript. The University of Alaska, National Science Foundation (Alaska EPSCoR), U.S. Forest Service (USFS) Yakutat Ranger District, USFS Tongass National Forest, USFS Alaska Regional Office, and USFS Pacific Northwest Research Station Juneau Forestry Sciences Laboratory provided support and/or project funding. Graduate fellowship funding for C.S. Shanley was provided by the US National Science Foundation IGERT Resilience and Adaptation
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Program (Award # 0654441) and the Alaska EPSCoR Program (Award #0701898). References ADFG (2008). Community Subsistence Information System (CSIS). Alaska Department of Fish and Game, Division of Subsistence, Juneau, Alaska. Retrieved from Accessed 24.05.12. Ford, J. D., Smit, B., & Wandel, J. (2006). Vulnerability to climate change in the Arctic: A case study from Arctic Bay, Canada. Global Environmental Change, 16, 145–160. Freeman, M. M. R. (1976). . Report of the Inuit land use and occupancy project Ottawa, Canada: Department of Indian and Northern Affairs. Gordon, A. B., Andre, M., Kaglik, B., Cockney, S., Allen, M., Snowshoe, A., et al. (2007). Arctic borderlands ecological knowledge co-op community reports 2005–2006. Borderlands Society, Whitehorse, Yukon. Retrieved from Accessed 24.05.12. Johnson, C. J., Nielsen, S. E., Merrill, E. H., McDonald, L. L., & Boyce, M. S. (2006). Resource selection functions based on use-availability data: Theoretical motivation and evaluation methods. Journal of Wildlife Management, 70, 347–357.
17
Kofinas, G. P., Chapin, F. S., III, BurnSilver, S., Schmidt, J. I., Fresco, N. L., Kielland, K., et al. (2010). Resilience of Athabascan subsistence systems to interior Alaska’s changing climate. Canadian Journal of Forest Research, 40, 1347–1359. Kruse, J. A., White, R. G., Epstein, H. E., Archie, B., Berman, M., Braund, S. R., et al. (2004). Modeling sustainability of arctic communities: an interdisciplinary collaboration of researchers and local knowledge holders. Ecosystems, 7, 815–828. Larsen, C. F., Motyka, R. J., Freymueller, J. T., Echelmeyer, K. A., & Ivins, E. R. (2005). Rapid viscoelastic uplift in southeast Alaska caused by post-Little Ice Age glacial retreat. Earth and Planetary Science Letters, 237, 548–560. Liu, C., Berry, P. M., Dawson, T. P., & Pearson, R. G. (2005). Selecting thresholds of occurrence in the prediction of species distributions. Ecography, 28, 385– 393. Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L., & Erickson, W. P. (2002). Resource selection by animals: Statistical design and analysis for field studies. The Netherlands: Kluwer Academic Publishers. Mills, D. D., & Firman, A. S. (1986). Fish and wildlife use patterns in Yakutat, Alaska: contemporary patterns and changes. Alaska Department of Fish and Game, Division of Subsistence, Douglas, Alaska. Retrieved from Accessed 24.05.12. Natcher, D. C. (2004). Implications of fire policy on Native land use in the Yukon Flats, Alaska. Human Ecology, 32, 421–441. NOAA (1983). Local climatological data annual summary Yakutat, Alaska. National Oceanic and Atmospheric Administration. U.S. Department of Commerce. Oehlers, S. A. (2007). Habitat Selection and Sightability of Moose in Southeast Alaska. MS Thesis. University of Alaska Fairbanks, Fairbanks, AK. Pedersen, S., & Coffing, M. (1984). Caribou hunting: land use dimensions and recent harvest patterns in Kaktovik, northeast Alaska. Alaska Department of Fish and Game, Division of Subsistence, Fairbanks, Alaska. Retrieved form Accessed 24.05.12. Peterson, G. D., Cumming, G. S., & Carpenter, S. R. (2003). Scenario planning: a tool for conservation in an uncertain world. Conservation Biology, 17, 358– 366. Peterson, G., Douglas Beard, T., Jr., Beisner, B. E., Bennett, E. L., Carpenter, S. R., Cumming, G. S., et al. (2003). Assessing future ecosystem services: a case study of the Northern Highlands Lake District, Wisconsin. Conservation Ecology, 7, Retrieved from Accessed 24.05.12. Shephard, M. E. (1995). Plant community ecology and classification of the Yakutat Foreland, Alaska. Juneau, Alaska: USDA Forest Service. Smith, C.A., & Franzmann, A.W. (1979). Productivity and physiology of Yakutat Forelands moose. Final Report W-17-10 and W-17-11. Alaska Department of Fish & Game. Sowl, K., & Poetter, R. (2004). Impact analysis of off-road vehicle use for subsistence purposes on refuge lands and resources adjacent to King Cove access project. U.S. Fish & Wildlife Service, Izembek National Wildlife Refuge, Alaska. Retrieved from Accessed 24.05.12. Stephenson, T. R., Van Ballenberghe, V., Peek, J. M., & MacCracken, J. G. (2006). Spatio-temporal constraints on moose habitat and carrying capacity in coastal Alaska: Vegetation succession and climate. Rangeland Ecology & Management, 59, 359–372. Tobias, T. N. (2000). Chief Kerry’s moose: a guidebook to land use and occupancy mapping, research design and data collection. Union of BC Indian Chiefs and Ecotrust Canada. Retrieved from Accessed 24.05.12. U.S. Census (1960). U.S. Census. Bureau of the Census. U.S. Department of Commerce. USDA Forest Service. (2007). Yakutat access travel management plan. Yakutat, AK: USDA Forest Service. USDA Forest Service. (2009a). Access and Travel Management on the Yakutat Ranger District: Environmental Assessment. Tongass National Forest R10-MB-619a. Retrieved from: Accessed 29.10.12. USDA Forest Service (2009b). Access and Travel Management on the Yakutat Ranger District: Decision Notice. Yakutat, AK. Retrieved from: Accessed 29.10.12. Walker, B., Carpenter, S.R., Anderies, J., Abel, N., Cumming, G., Janssen, M., et al. (2002). Resilience management in social-ecological systems: a working hypothesis for a participatory approach. Conservation Ecology, 6, Retrieved from Accessed 24.05.12. Walker, B., Holling, C. S., Carpenter, S. R., Kinzig, A. (2004). Resilience, adaptability and transformability in social-ecological systems. Ecology and Society, 9, Retrieved from
Contents lists available at SciVerse ScienceDirect
Landscape and Urban Planning journal homepage: www.elsevier.com/locate/landurbplan
Research paper
Balancing the conservation of wildlife habitat with subsistence hunting access: A geospatial-scenario planning framework Colin S. Shanley a,∗ , Gary P. Kofinas b , Sanjay Pyare c a b c
University of Alaska Fairbanks, Institute of Arctic Biology, Resilience and Adaptation Program, 902 N. Koyukuk Dr., P.O. Box 757000, Fairbanks, AK 99775, USA University of Alaska Fairbanks, Institute of Arctic Biology, Department of Humans and Environment, 902 N. Koyukuk Dr., P.O. Box 757000, Fairbanks, AK 99775, USA University of Alaska Southeast, Environmental Sciences and Geography Program, 11120 Glacier Highway, Juneau, AK 99801, USA
h i g h l i g h t s • • • •
Contemporary subsistence hunting increasingly relies on motorized access. Motorized access used for subsistence hunting may effect wildlife habitat. Habitat conservation and hunting access can be spatially integrated with GIS. Scenario planning can evaluate wildlife habitat and hunting access quantitatively.
a r t i c l e
i n f o
Article history: Received 20 June 2012 Received in revised form 25 January 2013 Accepted 11 March 2013 Available online 9 April 2013 Keywords: Access Adaptation Land-use mapping Resilience Scenario planning Transportation planning
a b s t r a c t Increased motorized access used for subsistence hunting has created a challenge for land managers trying to balance the conservation of wildlife habitat with the greater environmental impact of motorized access. We used an interdisciplinary approach to evaluate this challenge in a case study of subsistence moose (Alces alces) hunters who used off-highway vehicles (OHVs; e.g., four-wheelers) to access remote harvest areas in Yakutat, Alaska, USA, and the conservation needs to sustain moose. We developed a resiliencebased planning framework that combined methods from wildlife ecology, land-use mapping, and scenario planning. The study started at the community level by working with local hunters to evaluate their values and goals for subsistence moose hunting, and to identify thresholds of undesired change. This process served as the basis for evaluating how four road closure scenarios would effect the distribution of moose and hunters’ access to moose harvest areas. The effect of roads and OHV routes on moose distribution was quantified in a previous study with a GIS-based resource selection function model. An index of access was quantified on a digitized map of harvest areas. The results of the scenario analyses suggest that a balance in the conservation of wildlife habitat and subsistence hunting access could be found in the spatial arrangement of routes that are outside of important moose habitat, but within reach of preferred harvest areas. This spatially explicit planning framework may prove useful in northern communities experiencing an increased use of motorized access for contemporary subsistence hunting practices. © 2013 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Background Subsistence hunting communities across high latitude regions of the world are experiencing increased social, economic, and ecological changes that require rapid adaptation (Berkes & Jolly, 2001;
∗ Corresponding author. Present address: The Nature Conservancy, Alaska Field Office, 416 Harris Street, Suite 301, Juneau, AK 99801, USA. Tel.: +1 907 523 4929; fax: +1 907 586 8622. E-mail addresses: [email protected] (C.S. Shanley), gpkofi[email protected] (G.P. Kofinas), [email protected] (S. Pyare). 0169-2046/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.landurbplan.2013.03.006
Brinkman, Kofinas, Chapin, & Person, 2007; Condon, Collings, & Wenzel, 1995; Ford, Smit, & Wendel, 2006; Gordon et al., 2007; Kofinas et al., 2010). One such adaptation is the increased use of motorized access for more efficient hunting (e.g., four-wheelers and snowmachines). The increased use of motorized access for hunting has created new challenges for land management efforts that seek to balance the conservation of wildlife habitat with the need for subsistence access (Ahlstrand & Racine, 1993; Mills & Firman, 1986; Sowl & Poetter, 2004). The conservation of wildlife habitat and subsistence hunting with motorized vehicles are often considered mutually exclusive (Gordon et al., 2007; Mills & Firman, 1986; Sowl & Poetter, 2004). A challenge in interdisciplinary research on ecosystem stewardship is to develop new and creative approaches to meet both these
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
11
Fig. 1. A map of the study area, Yakutat, Alaska, USA and the network of existing vehicle routes. Yakutat is located along the Gulf of Alaska in the northernmost corner of the Tongass National Forest.
social and ecological goals (Chapin et al., 2006; Chapin, Kofinas, & Folke, 2009). For example, a resilience-based approach that uses road and trail closure scenarios could provide an effective framework to identify balanced solutions. Simultaneously considering both the social values of stakeholders and the ecological properties of a system increases the likelihood for more robust management decisions and resilience to unforeseen events (Berkes, Colding, & Folke, 2003; Chapin et al., 2009; Cordell & Bergstrom, 1999). Using a resilience-based approach, in abstract, one searches for ways to maintain valued attributes while avoiding thresholds to change in which valued attributes are lost (Walker et al., 2002). In this study, for instance, the valued attributes were identified as the conservation of wildlife habitat and access to subsistence resources. The ecological threshold was defined as the point in which wildlife habitat is effectively lost due to high levels of motorized disturbance (Shanley & Pyare, 2011). The associated social threshold is the point at which subsistence hunters do not have sufficient access to meet subsistence resources needs (Berman & Kofinas, 2004; Brinkman, Chapin, Kofinas, & Person, 2009). Scenario analysis can be used in an interdisciplinary manner to address resource management questions about future conditions, which ultimately builds resilience into the sustainability of valued ecosystem services (Ascher, 2009; Carpenter, Bennett, & Peterson, 2006; Peterson, Cumming, & Carpenter, 2003). In general, the objective is to generate an understanding of plausible (not necessarily likely) future conditions that provide the basis for discussions about the costs and benefits associated with alternative actions or policies (Carpenter et al., 2006). Scenario analysis has been undertaken to explore a wide variety of natural resource questions. For example, Peterson, Douglas Beard, et al. (2003) used a scenario planning approach on a regional scale in the Northern Highlands Lake District, Wisconsin, to explore the future consequences of urbanization and ecological vulnerability to undesired change, such as the loss of valued fishing opportunities. Focusing on the sustainability of arctic subsistence communities, Kruse et al. (2004) used scenario analysis to explore the multiple effects of climate change, oil development, and tourism on caribou hunting. Researchers in these and other projects have taken different approaches to scenario planning based on the central question, scope of the project, and availability of data.
To balance the conservation of moose habitat with off-highway vehicles (OHVs; e.g., four-wheelers), we developed spatially explicit scenarios to evaluate effects on both moose disturbance and subsistence hunting access. We used this geospatial approach for three reasons: (1) spatially explicit scenarios with GIS improves the evaluation and analysis of potential sources of disturbance on wildlife (Johnson, Nielsen, Merrill, McDonald, & Boyce, 2006); (2) many subsistence hunters have a strong spatial and visual orientation (Tobias, 2000); and (3) GIS allowed us to integrate and synthesize a variety of data types (i.e., wildlife distribution and land-use). The use of GIS technology for evaluating social and ecological data has become increasingly important for land-use planning. For example, Sawyer, Nielson, Lindzey, and McDonald (2006) used collar location data from mule deer (Odocoileus hemionus) and GIS to evaluate the trade-offs between deer habitat and natural gas development in Wyoming, U.S. Similarly, Braund (2007) used community interviews and participatory mapping with residents to create a geodatabase of traditional land-use near Tyonek and Beluga, Alaska, to evaluate potential industrial development conflict. However, new approaches are needed to combine these types of social and ecological analyses for more integrated assessments and adaptive management (Beier, Patterson, & Chapin, 2008). In this study, we illustrate how a resilience-based approach with spatially explicit scenario planning could provide an effective framework to achieve both the goals of habitat conservation and resource access. 1.2. Case study Our case study occurred on the Yakutat Ranger District of the Tongass National Forest in southeastern Alaska, USA (Fig. 1). The location and timing of subsistence hunting has traditionally been unregulated (Mills & Firman, 1986). With the advent of OHVs, subsistence hunters incorporated them into subsistence hunting strategies because OHVs allowed more ground to be covered in search of game and greater efficiency in transporting large game species from remote harvest areas (Mills & Firman, 1986). The continual use of OHVs has visibly affected the landscape with ruts in wetland areas that remain for years, even after a single event. The prevalence of such ruts became a concern for regional land
12
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Table 1 The study framework used to conduct an integrative geospatial-scenario analysis of subsistence hunting access and wildlife habitat conservation. 1. Identify values, goals, and thresholds to undesired change with respect to subsistence hunting access and wildlife habitat conservation. 2. Develop and map feasible multiuse access scenarios (e.g., road closures) with land managers using input and review from public hearings. 3. Interview resident subsistence hunters to map preferred harvest areas, elicit access constraints, and digitize results in a GIS. 4. Develop wildlife distribution models that include a variable(s) that quantifies the spatially explicit impact of access scenarios. 5. Quantitatively evaluate and rank the scenario’s effect on the extent of accessible harvest areas and the extent of high value wildlife habitat. 6. Determine if the results of any one scenario are acceptable for subsistence hunting access and wildlife habitat conservation, as well as any other criteria (e.g., ecosystem functions, cost, etc.). If not, apply lessons learned to develop and evaluate a new suite of alternative scenarios.
managers when anadromous (salmon) streams were affected and the wildlife species were potentially displaced due to disturbance (e.g., noise) (USDA Forest Service, 2007). Regulated access on the Tongass National Forest was mandated by the 1997 Tongass Land and Resources Management Plan (TLMP). Initial public meetings that introduced an Environmental Assessment revealed that residents were concerned that restricted motorized access would impede subsistence hunting practices. The subsistence activity of greatest concern with regards to access was the harvest of moose (Alces alces). Moose have historically been harvested in remote areas and these large animals (>500 kg) can be difficult to transport before the meat spoils. However, the large amount of meat forms a significant portion of a families’ diet, making as successful hunt important. Moose were also a species that land managers were concerned would be affected by unregulated access, given OHV damage had increased in areas thought to contain the region’s best moose habitat (USDA Forest Service, 2007). Our study objective was to develop and test a geospatialscenario planning framework (Table 1) with the goal of balancing the conservation of wildlife habitat and subsistence-hunting access. This study was conducted by evaluating the effect of four road-closure scenarios on both social and ecological goals of maintaining sufficient access and conserving moose habitat. Moose distribution models (Shanley & Pyare, 2011) that accounted for OHV impact on moose habitat were used to compare moose distribution among scenarios. Likewise, parameters of subsistence-hunting practices, elicited through moose hunter interviews, were extrapolated in GIS to compare the spatial extent of subsistence hunting access. 2. Methods 2.1. Social-ecological system Yakutat, Alaska, has approximately 800 residents with a mixed cash-subsistence economy. Approximately 60% of the residents are Alaska Native, primarily Tlingit. Tlingit subsistence activities were traditionally focused around marine resources with salmon as the main source of protein (Mills & Firman, 1986). During the early 1930s, moose migrated into the Yakutat area from interior Canada and were opportunistically harvested from river shores with fishing vessels to supplement marine protein sources. There was no evidence of moose in the area prior to that time. By the late 1980s, moose was the primary source of red meat with 70% of households consuming moose, either through direct harvest or sharing networks (Mills & Firman, 1986). In the 1960s, the Yakutat area was heavily vegetated with highquality moose forage (i.e., willow) and the moose population grew steadily (Mills & Firman, 1986). During this period, the moose
Fig. 2. The number of reported moose harvested in Yakutat, Alaska, USA from 1959 to 2007 (ADFG, 2008).
population size was believed to be >2000 (Smith & Franzmann, 1979) and the human population of Yakutat was approximately 300 residents with over 80% Alaska Native (U.S. Census, 1960). The road system was limited and hunting of moose was conducted primarily from fishing vessels along river outwash areas where a harvested moose could easily be transported. In the late 1960s, the road network expanded with logging activities, allowing increased hunting opportunities, new job opportunities and a concomitant increase in the human population (Mills & Firman, 1986). Competition also increased among residents of Yakutat and other communities (e.g., Juneau, Alaska), with hunters increasingly using OHVs to access remote harvest areas (Mills & Firman, 1986). The peak harvest of 380 moose occurred in 1970 (Fig. 2). Moose became a highly valued source of protein to indigenous and nonindigenous groups, and liberal harvest limits continued despite the belief of biologists that the moose population was declining (Mills & Firman, 1986). In the early 1970s, a series of severe winters with snowfalls >7 m (NOAA, 1983) resulted in high moose mortality levels. By the mid-1970s, the moose population declined to the point that hunting was closed between 1974 and 1977 (Fig. 2). Since then, the moose population has never returned to pre-1960s levels. Captured female moose in the late-1970s showed normal pregnancy rates, although they were shown to be nutritionally stressed (Smith & Franzmann, 1979). Long-term residents suggested the nutritional stress arose from the rapid change in the region’s vegetation composition, from higher quality moose-forage communities with willow (Salix spp.) and alder (Alnus sinuate) to lower quality communities like cottonwood (Populus trichocarpa) and spruce (Picea sitchensis) (Larsen, Motyka, Freymueller, Echelmeyer, & Ivins, 2005; Mills & Firman, 1986; Shephard, 1995). A similar pattern of successional change with an increase and decline in moose population was documented in an adjacent coastal area of the Copper River Delta, Alaska (Stephenson, Van Ballenberghe, Peek, & MacCracken, 2006). The moose population of Yakutat has remained relatively stable in recent history, utilizing early successional habitats created from natural disturbance events such as spring river flooding and avalanches, with approximately 800–1000 individuals estimated by aerial surveys during 2003 and 2004 (Oehlers, 2007). 2.2. Identifying values, goals, and thresholds In the spring of 2007, we held a series of informal meetings separately with land managers and resident moose hunters, with follow-up discussions for review of findings and clarification. The purpose of the meetings was to define each group’s respective
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
13
Table 2 Categorized route lengths (km) and sum for each of the four road closure scenarios in Yakutat, Alaska, USA. Routes (km)
Low OHV High OHV All-vehicles Total OHV All routes
Fig. 3. A ball-and-cup diagram illustrating the multiple stability states and threshold in which subsistence moose hunters can have sufficient – or insufficient – access to subsistence resources to meet their household needs in Yakutat, Alaska, USA.
values, management goals, and perceived thresholds to undesired changes with respect to moose conservation and harvesting access (e.g., Berkes, Mathias, Kislalioglu, & Fast, 2001). Land managers identified their goal as the management of habitat on the Yakutat landscape to support a healthy moose population. Due to the reduced population counts of moose on the portion of the region heavily used by OHVs, they suspected OHVs were effectively lowering the region’s carrying capacity for moose (USDA Forest Service, 2007). Managers were therefore primarily interested in whether or not there was a reduced probability of moose occurrence relative to OHV activity. Resident moose hunters emphasized the importance of using OHVs to retrieve large game species from remote harvest areas, since these species were critical to meeting their economic, nutritional, and cultural needs. Hunters indicated there was a maximum or threshold distance they were willing to travel by foot to transport a moose from a harvest site to an OHV route or road. This threshold distance can be illustrated with an adapted ball-and-cup diagram from the resilience literature (e.g., Walker, Holling, Carpenter, & Kinzig, 2004) that shows alternate stability states with more or less desirable outcomes (Fig. 3). Hunters also noted that this distance was dynamic: this study would be a “snapshot” in time that represents the current social and ecological conditions. For example, with new forms of hunting innovation and technology this distance would likely change. Nevertheless, moose hunters stated for the foreseeable future they were primarily interested in maintaining sufficient OHV access to important harvest areas. 2.3. Road closure scenarios Four road-closure scenarios under consideration by the Yakutat Ranger District to meet multiple objectives (e.g., wildlife values, stream restoration, and maintenance costs) were used for comparative analyses (USDA Forest Service, 2007; Table 2). Scenario 1 maintained the status quo with unlimited OHV access. All main roads (221 km) and the extensive network of hunter-created OHV routes (302 km) across inland meadows and beaches would remain available to search for moose and transport a harvested moose back to the community. Scenario 2 retained most of the main roads (182 km) and restricted many of the hunter created OHV
Scenarios I
II
III
IV
184 118 221 302 523
54 31 182 85 267
18 21 185 39 224
21 34 182 55 237
routes to a select few (85 km) routes that could be maintained and monitored to National Forest standards (Table 2). A seasonal closure would also be put on OHV routes near tern (Onychoprion aleuticus and Sterna paradisaea) nesting habitat (May to mid-August). Hunters would be allowed to use an OHV to retrieve a harvested moose from open roads and OHV routes by obtaining a permit from the Yakutat Ranger District prior to hunting, and if they could do so without causing resource damage. Resource damage was defined as “soil displacement or cutting of living vegetation to create a path” from the road or OHV route to a harvested moose (USDA Forest Service, 2007). Hunters would also have to attain a permit from the Alaska Department of Natural Resources to cross streams used by anadromous salmon. Scenario 3 was similar to Scenario 2 in terms of the total amount of main roads (185 km), however subsistence hunting access was not given priority over national transportation planning standards (Table 2). Therefore, Scenario 3 retained fewer OHV routes than any of the other scenarios (39 km). Perhaps most importantly, hunters would also not be able to use OHVs to retrieve moose from open roads or OHV routes. Finally, Scenario 4 incorporated public comments on Scenario 2 from a meeting with the Yakutat Ranger District in 2005. This scenario resulted in fewer main roads (182 km), many of which lacked utility (e.g., old logging roads), and a greater number of designated OHV routes (55 km) (Table 2). Hunters in Scenario 4 were also allowed to use OHVs to transport a harvested moose to open roads and OHV routes, if they could do so without causing resource damage and obtain a permit prior to hunting. Seasonal closure in areas of nesting terns and a permitting process for crossing of salmon streams would also apply to Scenario 4. 2.4. Evaluating subsistence access A random sample of one-third (n = 25) of the federally registered subsistence moose hunters were interviewed to document their land-use patterns in 2007. During these 45-min interviews, hunters (n = 21 of 25) mapped their preferred harvest areas and estimated (n = 25) how far they were willing to transport a harvested moose on foot to an OHV route or road. A preferred harvest area was defined as an area currently valued for hunting moose. Each hunter delineated harvest areas with dry erase markers on a transparency over a 1.5 m x 1 m high-resolution aerial photograph of the region (approximately 2 cm = 1 km). Harvest areas for each hunter were manually digitized into GIS in the form of grids of cells (50 m resolution) designated with 1 of 2 values: 1 for harvest and 0 non-harvest. These 21 grids were collectively summed to create one grid for the sampled hunter population. For example, if the harvest areas of three hunters overlapped, cells in the composite harvest surface acquired a value of three, indicating that 14% (3 hunters/21 total hunters) used the cell area for hunting. The distance that each hunter was willing to transport a harvested moose without motorized assistance to a road or OHV route was elicited during interviews. The mean (1.5 km), lower 95% CI (0.6 km), and upper 95% CI (2.4 km) transport distances were used for comparative analysis to represent the range of hunting
14
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Table 3 Overlapping harvest area (km2 ) categories and access score in the four road closure scenarios for Yakutat, Alaska, USA. The access score was derived from multiplying the km2 of access by the percent use category, and then summing across categories for each scenario and distance. The hunter access distances are 0.6 km (“Short”), 1.5 km (“Average”), and 2.4 km (“Long”). Scenarios Hunter harvest area percent use category
I
II
Distance
5% 10% 14% 19% 24% 29% 33% Access score
III
Distance
IV
Distance
Distance
Short
Average
Long
Short
Average
Long
Short
Average
Long
Short
Average
Long
272.5 206.7 76.9 37.0 10.8 3.5 0.3 54.4
423.6 325.8 144.2 72.9 17.1 3.5 0.3 90.8
495.1 379.6 196.0 96.8 22.2 4.2 0.3 112.7
164.4 68.5 16.8 10.0 4.2 0.0 0.0 19.7
274.0 151.0 41.7 17.0 6.8 0.5 0.0 38.4
346.7 226.8 66.0 29.2 10.0 1.4 0.0 55.9
162.3 56.2 12.4 4.0 2.2 0.0 0.0 16.1
259.3 126.0 33.4 8.1 4.3 0.0 0.0 31.7
319.8 197.5 50.5 12.8 6.4 0.8 0.0 45.4
171.3 72.6 18.3 10.1 4.2 0.0 0.0 20.6
294.6 161.5 50.0 17.4 6.8 0.5 0.0 41.6
383.2 243.1 80.4 31.9 10.0 1.4 0.0 61.7
zones around roads and OHV routes in each scenario. These zones represented portions of the landscape that were considered accessible by the of resident-hunter population. An access score was then derived for each scenario and transport distance by overlaying the scenario-specific access zones on the cumulative harvest grid. First, harvest grid-cell values occurring within access zones for each road closure scenario were summed for each percent-use category. Then, we assumed a linear relationship between the relative importance of cells and the percent of hunters using the hunting area on the cumulative harvest grid. This assumption was made to weight the relative importance of harvest areas to the subsistence hunting community as a whole for scenario analysis. To calculate each scenario’s final access score, the total cell area of each percent-use category was multiplied by the percentage of hunters that used the area those cells represented, and summed across all percent-use categories.
2.5. Evaluating moose habitat A comparison of the cumulative impact of road-closure scenarios on the amount of high-probability female moose habitat (Shanley & Pyare, 2011) was developed using a cumulative habitat score for each scenario. To do this, we used GIS to map the proposed configuration of roads and OHV routes for each of the four scenarios. Each route was categorized as either Low OHV, High OHV, and All-Vehicles by size and surface wear, and then weighted with the mean number of seasonal one-way trips for each route category (Shanley & Pyare, 2011); a parameter estimated from hunter interviews. A female moose distribution model (Shanley & Pyare, 2011) was used to evaluate the effect of the four road closure scenarios on moose distribution. A female distribution model was selected due to their reproductive importance and higher likelihood of disturbance (Shanley & Pyare, 2011). The model was a spatially explicit resource selection function (RSF Design II; Johnson et al., 2006; Manly, McDonald, Thomas, McDonald, & Erickson, 2002) that incorporated a parameter for OHV activity derived from interviews, as well as habitat composition and configuration. The resulting RSF for each scenario was a mapped surface of 20 m × 20 m grid cells representing the probability of moose occurrence across the study area. We utilized an average predicted probability/suitability approach (Liu, Berry, Dawson, & Pearson, 2005) where the mean probability of occurrence was used to determine a 0.4 threshold probability value for delineating species occurrence across the study area. Cells above the threshold probability value were classified into 0.1 probability intervals (e.g., 0.4–0.5, 0.5–0.6, etc.) for scenario analysis. We assumed a linear relationship between the probability of moose
Table 4 Moose habitat (km2 ) by probability interval and habitat score in the four road closure scenarios for Yakutat, Alaska, USA. The habitat score was derived by multiplying the km2 of habitat by the probability of use interval mid-point, and then summing across probability-corrected intervals for each scenario. Moose habitat probability intervals
0.4–0.5 0.5–0.6 0.6–0.7 0.7–0.8 0.8–0.9 0.9–1.0 Habitat score
Scenarios
I
II
III
IV
231.4 196.0 72.9 23.3 7.9 0.2 283.7
263.5 218.6 75.4 23.4 7.9 0.2 312.3
263.9 222.4 75.7 23.4 7.9 0.2 314.7
262.8 218.6 75.4 23.4 7.9 0.2 312.0
occurrence and the relative importance of habitat, and the total area of habitat in each probability interval was multiplied by the average probability of the interval. This assumption was made to weight the relative importance of habitats for scenario analysis. The final habitat score was calculated from the sum of the probability-corrected habitat intervals for each road closure scenario. 3. Scenario results Scenario 1 maintained the status quo and included the greatest number of roads and OHV routes; and consistently resulted in the best scores for subsistence hunting access across the range of short, average, and long-distance hunter categories (Table 3 and Fig. 4). It also resulted in the lowest score for female moose habitat (Table 4 and Fig. 5), primarily seen in the 0.4–0.6 probability categories, because female moose were influenced by this larger road network. Scenario 2 (the U.S. Forest Service-preferred alternative) resulted in the second lowest scores for subsistence hunting access (Table 3) and a habitat score for female moose similar to Scenario 4 (Table 4). The closing of many roads and OHV routes would approximately halve the access scores of Scenario 1, but result in greater habitat area for female moose. Scenario 3, which met minimum policy requirements, resulted in the lowest score for subsistence hunting access (Table 3) and the greatest habitat score for female moose (Table 4). The low access scores and the high habitat score were the result of fewer roads and OHV routes. Scenario 4 was based on public comments on Scenario 2 and yielded the second highest scores for subsistence hunting access (Table 3) and a habitat score for female moose similar to Scenario 2 (Table 4). Scenario 4 had fewer roads but more OHV routes, which could explain why it provided for greater hunter access (Fig. 4).
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
15
Fig. 4. Map of subsistence moose hunters’ preferred harvest areas by percent use categories in Yakutat, Alaska, USA and the average distance (hashed areas) hunters are willing to transport a harvested moose from the existing network of vehicle routes in Scenario 1. Increasingly darker harvest areas represent increased hunter use. An inset (a) shows an area of extensive OHV routes in Scenario 1 and (b) shows the same visual extent and access to preferred harvest areas with fewer OHV routes in Scenario 4.
4. Discussion The results of the road closure scenarios illustrate a socialecological dynamic created by the spatial configuration of routes on moose habitat and subsistence hunting access. Comparison of the results of Scenario 1 with Scenario 4 illustrates how public comments incorporated in Scenario 4 consistently improved hunting access for short, average, and long-distance hunter categories (Table 3) with minimal impact on moose habitat (Table 4). The existing network in Scenario 1 yielded a greater degree of access but also had a noticeable disturbance effect on female moose (Table 4
and Fig. 5). Scenario 4 had fewer total routes but more harvest areas within reach of roads and OHV routes than any of the other scenarios. Maintaining the status quo in Scenario 1 without enacting new regulations would also allow future OHV route expansion that could pose a future impact to female moose habitat. The proposed regulations for Scenario 4 would ban the creation of new OHV routes but allow the retrieval of a harvested moose from designated routes, therefore limiting future impact on female moose habitat and allowing for more efficient transportation of a harvested moose. Additionally, the closure of main roads (All-Vehicle routes) seen in Scenario 1 illustrates how, kilometer-for-kilometer, the
Fig. 5. Resource selection function model of female moose distribution with the existing network of vehicle routes in Scenario 1; illustrating the impact of roads and OHV routes on female moose habitat selection in Yakutat, Alaska, USA. Increasingly darker areas represent a higher probability of moose occurrence. An inset (a) shows an area of extensive OHV routes in Scenario 1 and (b) shows the same visual extent with fewer OHV routes in Scenario 4.
16
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
closure of main roads would have resulted in greater habitat gains for moose than the closure of OHV routes (Fig. 5) due to the higher disturbance effect caused by main roads (Shanley & Pyare, 2011). Scenario 4 was ultimately selected by the Yakutat Ranger District as the adopted plan (USDA Forest Service, 2009a, 2009b). Although Scenario 4 showed promise through the incorporation of public comments, no clear “winner” emerged as the best scenario for the conservation of moose habitat and subsistence hunting access. Indeed, it may be too simplistic to utilize a linear relationship to score the extent of moose habitat and hunting access as proxies for evaluating planning scenarios. Rather, a balance in the conservation of wildlife habitat and subsistence hunting access may be achieved via the precise spatial configuration of transportation networks. More specifically, a balance could be achieved by the strategic closure of roads and OHV routes in specific areas with high probability moose habitat and the maintenance of routes that are within reach of important harvest areas. The methodology demonstrated here – the spatial integration of ecological and subsistence access – could likewise be applied to finer-scale scenario planning involving individual routes and specific harvest areas. Follow-up interviews with the Yakutat Ranger District in 2012 suggest the implementation of Scenario 4 has had successes and challenges that will likely require future amendments and finerscale analysis. While Scenario 4 has allowed for improved habitat protection in closed areas, the provision to allow OHV use for moose harvest retrieval has presented a challenge for assessing fine-scale environmental impacts adjacent to open roads and OHV routes. New studies are underway to quantify the environmental impacts using a system of hunter reporting and harvest site surveys. The land managers suggested the results of these studies will likely result in additional OHV route closings and in turn new trail openings to achieve a balance in the conservation of wildlife habitat and subsistence hunting access. The results of the scenario analyses to date should be treated within the context of the assumptions made in our spatial analyses. The distance hunters were willing to transport a harvested moose to an OHV route or road was clearly identified as an important management consideration, but simplifying it to a range of three distances for our analyses should not overshadow the complexity of this metric that depends on yearly hunting conditions (e.g., weather), and will likely change as hunters adapt to new social and ecological conditions. Similarly, the disturbance of female moose will depend on the spatial configuration of habitats (Shanley & Pyare, 2011) that will also likely change with continued successional processes (Larsen et al., 2005; Shephard, 1995; Stephenson et al., 2006). Nevertheless, the resilience-based approach using a geospatial-scenario framework provided a quantitative and interdisciplinary process for assessing future management iterations. Indeed, increased social and ecological change, e.g., climate change, will necessitate flexibility as a crucial management consideration in the development of regulations that manage subsistence hunting access. A review of other land-use studies suggests a geospatialscenario framework may prove useful in many other regions where motorized subsistence access is increasing and the conservation of wildlife habitat is a concern. Some of the first subsistence landuse studies occurred in the Canadian Arctic, documenting Native land claims in the face of industrial development (Freeman, 1976). The results of these studies demonstrated how subsistence mapping could elicit detailed use of the landscape and the distance hunters traveled to reach harvest areas. Subsequent studies have also produced species-specific harvest maps, demonstrating highvalue harvest areas that have stayed in the same locations for generations or alternatively, where social and ecological conditions have changed and required adaptation through, for instance, motorized vehicles to hunt more efficiently (Berkes et al., 1995;
Natcher, 2004; Pedersen & Coffing, 1984). The variety of subsistence land-use patterns depends on the region and species of interest, and underscores the importance of time and place specific analyses. For example, Berman and Kofinas (2004) showed that the subsistence caribou hunters of Old Crow, Yukon preferred to harvest caribou on time-tested accessible migration routes in order to save resources (i.e., time and money), when possible. When the migrations changed, additional resources had to be used on extended trips with motorboats and snowmachines to secure a sufficient harvest. A similar pattern of harvest in local areas mixed with long-distance motorized travel was shown by Natcher (2004) in the Yukon Flat communities of interior Alaska. Moose hunters of Birch Creek preferred to hunt local wetland areas, but changing interior fire regimes (natural and prescribed) caused a time lag in moose-forage availability between burns in these wetland areas. The changing distribution of moose required an additional investment of time and money to search for moose with motorboats or snowmachines. In addition to the variety of land-use patterns described in subsistence studies, it is clear that accessing and securing subsistence resources is changing, and motorized access is shaping how hunters perceive and use the landscape (Brinkman et al., 2009).
5. Conclusion The use of motorized access for the harvest of subsistence resources is likely to increase in Yakutat (USDA Forest Service, 2007), as it is in many northern subsistence communities (Berkes et al., 1995; Condon et al., 1995; Ford et al., 2006; Gordon et al., 2007). Increased integration of subsistence harvesting practices with the cash economy is likely to mean contemporary subsistence hunters will have less time to spend on the land than past generations, requiring more efficient means of accessing hunting areas. In addition, the expense of modern equipment such as snowmachines, OHVs, rifles, ammunition, and fuel, will result in an even greater involvement in the cash economy (Berman & Kofinas, 2004; Fast & Berkes, 1998). Furthermore, shifts in regional climate regimes will compound social-ecological changes due to greater investment in locating and securing the shifting distribution of resources (Chapin et al., 2004). In some regions or with some species, the influence of motorized access on wildlife habitat may be negligible (i.e., highly dispersed wildlife) or ecological changes will improve hunting opportunities. However, as this study suggests, access management plans in some regions need to be updated to better reflect changing social and ecological conditions. A resilience-based approach that incorporates geospatial-scenarios with integrative quantitative and qualitative components may provide an effective framework to balance the conservation of wildlife habitat with contemporary subsistence hunting practices.
Acknowledgements We would like to thank the hunters who participated in this study for sharing their time, knowledge, and hunting experiences. Members of the Yakutat Tlingit Tribe J. Ramos and E. Henninger, and B. Lucey of the Yakutat Salmon Board provided us with immense local insight. A.L. Jacob made helpful comments on the developing manuscript. The University of Alaska, National Science Foundation (Alaska EPSCoR), U.S. Forest Service (USFS) Yakutat Ranger District, USFS Tongass National Forest, USFS Alaska Regional Office, and USFS Pacific Northwest Research Station Juneau Forestry Sciences Laboratory provided support and/or project funding. Graduate fellowship funding for C.S. Shanley was provided by the US National Science Foundation IGERT Resilience and Adaptation
C.S. Shanley et al. / Landscape and Urban Planning 115 (2013) 10–17
Program (Award # 0654441) and the Alaska EPSCoR Program (Award #0701898). References ADFG (2008). Community Subsistence Information System (CSIS). Alaska Department of Fish and Game, Division of Subsistence, Juneau, Alaska. Retrieved from
17
Kofinas, G. P., Chapin, F. S., III, BurnSilver, S., Schmidt, J. I., Fresco, N. L., Kielland, K., et al. (2010). Resilience of Athabascan subsistence systems to interior Alaska’s changing climate. Canadian Journal of Forest Research, 40, 1347–1359. Kruse, J. A., White, R. G., Epstein, H. E., Archie, B., Berman, M., Braund, S. R., et al. (2004). Modeling sustainability of arctic communities: an interdisciplinary collaboration of researchers and local knowledge holders. Ecosystems, 7, 815–828. Larsen, C. F., Motyka, R. J., Freymueller, J. T., Echelmeyer, K. A., & Ivins, E. R. (2005). Rapid viscoelastic uplift in southeast Alaska caused by post-Little Ice Age glacial retreat. Earth and Planetary Science Letters, 237, 548–560. Liu, C., Berry, P. M., Dawson, T. P., & Pearson, R. G. (2005). Selecting thresholds of occurrence in the prediction of species distributions. Ecography, 28, 385– 393. Manly, B. F. J., McDonald, L. L., Thomas, D. L., McDonald, T. L., & Erickson, W. P. (2002). Resource selection by animals: Statistical design and analysis for field studies. The Netherlands: Kluwer Academic Publishers. Mills, D. D., & Firman, A. S. (1986). Fish and wildlife use patterns in Yakutat, Alaska: contemporary patterns and changes. Alaska Department of Fish and Game, Division of Subsistence, Douglas, Alaska. Retrieved from