SPECIES STATUS ASSESSMENT for the ALEXANDER ARCHIPELAGO WOLF (Canis lupus ligoni) U.S. Fish and Wildlife Service Region 7, Alaska November 23, 2015 Suggested citation: U.S. Fish and Wildlife Service. 2015. Species status assessment for the Alexander Archipelago wolf (Canis lupus ligoni). Version 1.0, December 2015. Alaska Region, Anchorage, Alaska. 162 pp. Species Status Assessment for the Alexander Archipelago Wolf EXECUTIVE SUMMARY The purpose of this document is to synthesize scientific information relevant to assessing the current and future status of the Alexander Archipelago wolf (Canis lupus ligoni). It will be used to inform the U.S. Fish and Wildlife Service’s (Service) decision on whether or not the Alexander Archipelago wolf warrants protection under the U.S. Endangered Species Act (Act). As such, it is not a decision document; instead, it forms the scientific basis from which the Service will draw conclusions and make decisions regarding the Alexander Archipelago wolf under the Act. In this assessment, we review the morphology, taxonomy, distribution, life history, ecology, and population dynamics of the Alexander Archipelago wolf and, as appropriate, its primary prey, black-tailed deer (Odocoileus hemionus spp.). We acknowledge uncertainty surrounding the taxonomic status of the wolf, but after careful review of the best available information, we assume for the purpose of this assessment that the Alexander Archipelago wolf is a valid subspecies of gray wolf (C. lupus). We then describe current habitat and resource conditions of the Alexander Archipelago wolf and project future conditions by evaluating effects of environmental and anthropogenic stressors to wolves at the individual, population, and rangewide levels. We conclude this assessment by characterizing future status of the Alexander Archipelago wolf using the conservation biology principles of redundancy (ability to withstand catastrophic events), resiliency (ability to withstand stochastic disturbance events), and representation (ability to adapt to changing environmental conditions). The Alexander Archipelago wolf occurs on the mainland of southeastern Alaska and coastal British Columbia west of the Coast Mountain Range and on larger islands except Admiralty, Baranof, and Chichagof islands and all of the Haida Gwaii, or Queen Charlotte, Islands (Figure ES-1). Its current range, which is similar to its recent historical range, covers roughly 217,000 km2. These coastal wolves probably interact at low levels with interior, continental wolves via trans-boundary rivers and low mountain passes. Therefore, distribution boundaries of the Alexander Archipelago wolf are porous and are not defined sharply or with certainty; zones of intergradation with interior, continental wolves exist. We estimated the rangewide population of the Alexander Archipelago wolf to be approximately 850–2,700 wolves with the majority (~62%) occurring in coastal British Columbia where populations have been stable since 2000. In southeastern Alaska (38% of the rangewide population), trend is estimated for only the population occupying Prince of Wales Island and surrounding islands (6% of rangewide population), which constitutes Game Management Unit (GMU) 2. Between 1994 and 2014, the GMU 2 wolf population declined by about 75% (SE=15), although confidence intervals of the point estimates overlap. For the remainder of southeastern Alaska (32% of rangewide population), population trend is not known. Generally, populations of Alexander Archipelago wolf are connected to one another, although some geographical disruptions exist due to the island geography within its range; the GMU 2 population is the most insular population. The Alexander Archipelago wolf appears to be a habitat and diet generalist, although it exhibits some general preferences. These coastal wolves spend most of their time at elevations below 400 ii Species Status Assessment for the Alexander Archipelago Wolf m, probably because abundance of prey typically is higher at low elevations compared to higher elevations. Their diet is highly variable across the range and seasons, but similar to gray wolves, ungulates compose a large portion of it with deer being the most common ungulate species available. The GMU 2 wolf population is more dependent on deer as prey compared to other coastal wolf populations because deer are the only ungulate available in GMU 2; elsewhere, at least two additional ungulate species occur. Consistent with their opportunistic predatory behavior, Alexander Archipelago wolves also consume marine and intertidal species including salmon (Oncorhynchus spp.) and marine mammals when and where available. We identified multiple stressors that may be impacting individuals and populations of the Alexander Archipelago wolf, although most of them have the potential to affect wolves indirectly, not directly. Key stressors examined as part of this assessment include timber harvest, road development, wolf harvest, and climate-related events. Of these, wolf harvest is the only source of direct mortality that may have an impact at the population and rangewide levels. Although road development has little direct effect on wolves, roads provide access for hunters and trappers to areas that otherwise may be inaccessible or difficult to access. Timber harvest and winter severity influence deer habitat capability and abundance, which can impact wolf populations, especially if other ungulate species are not available. We also considered a variety of other stressors such as effects of small populations, oil development, overexploitation of salmon, and hybridization with dogs. Because many stressors that may be affecting Alexander Archipelago wolves interact with one another, sometimes synergistically, we revised an existing, but outdated model of a hypothetical wolf population in GMU 2, to help determine the relative strength of influence of each stressor and the cumulative impact on wolves. In the model, we explicitly considered timber harvest, frequency of severe winters, and wolf harvest as functions of road development and ocean distance from towns and villages (both measures of access for hunters and trappers). We also used the model to predict wolf population trajectory in GMU 2, the area for which the most data on wolf population dynamics exist, under six scenarios representing possible future conditions. We summarize results of the model in this status assessment and fully describe the model, its assumptions, and outputs in Gilbert et al. (2015). We then evaluated the relative level of resiliency (low, intermediate, high) of five populations, or group of populations, of the Alexander Archipelago wolf by examining the magnitude of stressors and their known or expected effect on wolves. Those populations included southern coastal British Columbia (Regions 1 and 2), northern coastal British Columbia (Regions 5 and 6), mainland southeastern Alaska (GMUs 1 and 5A), and GMUs 2 and 3 in southeastern Alaska (Figure ES-1). For three populations (i.e., southern and northern coastal British Columbia and GMU 2), we relied on trend information to inform our assessment of their resiliency to stressors, individually and cumulatively. However, for two populations in southeastern Alaska (GMUs 1 and 5A, and GMU 3), we lacked trend information. Therefore, we compared magnitude of stressors to those populations with those of populations for which trend information exists and then assigned a level of resiliency based on degrees of similarity and difference; for these two populations, uncertainty regarding resiliency and future status is greater than for the three populations with existing trend information. iii Species Status Assessment for the Alexander Archipelago Wolf Figure ES-1. Assumed range of the Alexander Archipelago wolf (C. l. ligoni), and level of resiliency of individual or groups of populations, which were defined using boundaries of Game Management Units (GMU) in southeastern Alaska and Regions in coastal British Columbia. iv Species Status Assessment for the Alexander Archipelago Wolf Of the five populations evaluated, we found that three of them, composing 80% of the rangewide population, exhibit high resilience to stressors (Figure ES-1). Both populations in coastal British Columbia have been stable since 2000 despite intensive and extensive timber harvest in the southern portion (Regions 1 and 2; 30% of the forest logged) and in the northern portion (Regions 5 and 6; 16% of the forest logged). We attribute their resiliency to the availability of ungulate species other than deer as prey and to apparently sustainable rates of wolf harvest (average reported harvest of <7% of the population annually). Further, these populations likely encounter few disruptions to demographic and genetic connectivity, although we found no estimates of dispersal specific to wolves in coastal British Columbia. Based on similarities in the overall magnitude of stressors and population characteristics, such as ungulate prey availability, with coastal British Columbia, we determined that the Alexander Archipelago wolf population on the mainland of southeastern Alaska (GMUs 1 and 5A) also exhibits high resilience, although we lack trend estimates for this population and therefore are less certain of its ability to withstand stochastic disturbances. The GMU 2 wolf population, which constitutes 6% of the rangewide population, demonstrates low resilience to stressors, specifically the synergistic effects of wolf harvest and timber harvest (Figure ES-1). Although this population appears to be harvested at sustainable rates (average of 17% of the population annually), unreported harvest contributes substantially to total wolf harvest (38–45% of total harvest) in GMU 2, resulting in unsustainable rates of wolf harvest in some years. High rates of total harvest in GMU 2 have been facilitated by the highest levels of road and boat access for hunters and trappers across the range of the taxon. In addition, approximately 23% of the forest has been logged, likely reducing numbers of deer, the only ungulate species available as prey. The combination of these factors likely has caused an apparent population decline of about 75% (SE=15) since 1994 and, as predicted by our population model, wolf abundance in GMU 2 is expected to decline by another roughly 8–14% of current levels over the next 30 years. Lastly, we determined that the GMU 3 wolf population in central southeastern Alaska (14% of the rangewide population), shows an intermediate level of resiliency to stressors (Figure ES-1). However, like the population on the mainland of southeastern Alaska (i.e., GMUs 1 and 5A), we lack a trend estimate for the GMU 3 population and therefore, we used a comparative approach. The GMU 3 wolf population has similarities with both a stable population in northern coastal British Columbia (Regions 5 and 6; e.g., level of timber harvest) and with a declining population in GMU 2 (e.g., island geography). For example, in GMU 3, 14% of the forest has been logged, reducing deer habitat capability, although wolves have access to ungulate prey other than deer; these attributes are similar to those in northern coastal British Columbia. Yet, rates of reported harvest in GMU 3 (21% of population annually) are slightly higher than those in GMU 2 (17% of population annually), although we found no evidence indicating that unreported harvest in GMU 3 is occurring at or near the high rates documented in GMU 2; in addition, road and boat access for hunters and trappers in GMU 3 is lower than that in GMU 2. Thus, in considering the evidence collectively, we classified the GMU 3 population as exhibiting an intermediate level of resiliency, in part owing to its island geography. In conclusion, we believe that the future status of the rangewide population of the Alexander Archipelago wolf likely will be stable or perhaps slightly lower than current levels based on its v Species Status Assessment for the Alexander Archipelago Wolf resiliency, redundancy, and representation. We found that (1) most (80%) of the rangewide population exhibits high resilience to stochastic disturbance events; (2) multiple populations are distributed across a broad range, demonstrating redundancy for withstanding catastrophic events, although two island populations (i.e., GMUs 2 and 3) constituting 20% of the rangewide population are more insular than the mainland populations; and, (3) as a habitat and diet generalist, the Alexander Archipelago wolf exhibits a high degree of ecological diversity and most populations appear to harbor sufficient levels of genetic diversity with no evidence of genetic bottlenecking; both of these characteristics indicate representation, or the ability to adapt to changing environmental conditions. Owing to predicted declines in the GMU 2 wolf population, it is likely that the rangewide population will decrease in the future, but we expect the overall effect to be minor given that the GMU 2 population constitutes only 6% of the rangewide population, is geographically peripheral to the other populations, and appears to serve as a sink population. Nonetheless, the persistence of the GMU 2 population is desired and requires careful management actions and decisions to ensure its future health. vi Species Status Assessment for the Alexander Archipelago Wolf TABLE OF CONTENTS Executive summary ii Table of contents vii List of figures ix List of tables xi 1.1. Purpose and focus of this assessment 1 1.2. Geographic scope 2 Chapter 1: Introduction 1.3. Review of previous efforts 4 1.4. Terminology used in this assessment 4 2.1. Physical description 6 2.2. Taxonomy 8 2.2.1. Morphometric analyses 8 Chapter 2: Description of the 2.2.2. Genetic analyses 9 Alexander Archipelago wolf 2.2.3. Other relevant analyses 13 2.2.4. Uncertainty of taxonomic status 14 2.3. Distribution 15 3.1. Vital rates 17 3.1.1. Abundance and trend 17 3.1.2. Reproduction 22 3.1.3. Survival 22 Chapter 3: Life history and 3.1.4. Within-population dispersal 23 ecology 3.1.5. Sex ratio 24 3.2. Social organization 25 3.3. Ecology 25 3.3.1. Food habits 26 3.3.2. Space and habitat use 32 4.1. Connectivity 39 Chapter 4: Dynamics of multiple 4.1.1. Demographic connectivity 39 populations 4.1.2. Genetic connectivity 40 4.2. Population processes 42 5.1. Environment 43 5.2. Land ownership 44 5.3. Cause and effect analysis 47 5.3.1. Wolf population model 47 5.3.2. Timber harvest 48 5.3.3. Road development 68 Chapter 5: Current and future 5.3.4. Wolf harvest 75 habitat and resource conditions 5.3.5. Disease 95 5.3.6. Climate-related events 98 5.3.7. Other 102 5.3.8. Summary of stressors 104 5.4. Existing conservation mechanisms 106 5.4.1. Southeastern Alaska 106 5.4.2. Coastal British Columbia 108 6.1. Biological considerations 110 Chapter 6: Current and future 6.2. Assessment by population 112 status of the Alexander 6.2.1. GMU 2 wolf population 114 Archipelago wolf 6.2.2. GMUs 1 and 5A wolf population 118 vii Species Status Assessment for the Alexander Archipelago Wolf 6.2.3. GMU 3 wolf population 119 6.2.4. Region 5/6 wolf population 120 6.2.5. Region 1/2 wolf population 121 6.2.6. Summary of individual populations 122 6.3. Characterizing future status 124 6.3.1. Redundancy 124 6.3.2. Resiliency 124 6.3.3. Representation 124 6.4. Summary 127 Literature cited 128 Appendix I: Land area (km2) of Game Management Units and Regions 150 Note on typography: This document consists of six chapters, a list of literature cited, and two appendices. Chapters are presented in all capital letters (primary headings); sections are labeled by chapter then section number (e.g., 3.1 refers to Chapter 3, Section 1) and are underlined (secondary headings); and, subsections are denoted by chapter, section number, and subsection number (e.g., 3.1.2. refers to Chapter 3, Section 1, Subsection 2) and are presented in bold typeface (tertiary headings). Primary, secondary, and tertiary headings are listed in the Table of Contents. In some subsections, we found it necessary to include fourth- and fifth-order headings. Fourth-order headings are labeled in italics and underlined and fifth-order headings are presented in standard text; in both cases, the heading is followed by an Em Dash (i.e., “—“) and the paragraph begins on the same line. Tables and figures are numbered in chronological order as they appear in the text. Typography example: CHAPTER Section Subsection Fourth-order heading.—Paragraph begins. Fifth-order heading.—Paragraph begins. viii Species Status Assessment for the Alexander Archipelago Wolf LIST OF FIGURES Figure ES-1. Assumed range of the Alexander Archipelago wolf (C. l. ligoni), and level of resiliency of individual or groups of populations, which were defined using boundaries of iv Game Management Units in southeastern Alaska and Regions in coastal British Columbia. Figure 1. Assumed range of the Alexander Archipelago wolf (C. l. ligoni), as reviewed in this 3 assessment, southeastern Alaska and coastal British Columbia. Figure 2. Range of the Alexander Archipelago wolf by Game Management Unit in the 7 southern portion of southeastern Alaska. Figure 3. Estimates of Φ and Φ from three studies of mitochondrial DNA variation in ST CT 12 coastal and inland wolves from northwest North America. Figure 4. Game Management Unit boundaries in southeastern Alaska and Region boundaries 16 in British Columbia that are within the assumed range of the Alexander Archipelago wolf. Figure 5. Land cover across southeastern Alaska south of Yakutat. 44 Figure 6. Map depicting land ownership and management across southeastern Alaska south 46 of Yakutat. Figure 7. Age distribution of logged forest across all land ownerships in southeastern Alaska. 51 Figure 8. Current distribution of (unlogged) productive old-growth forest and (logged) 53 young-growth forest across southeastern Alaska with Game Management Unit boundaries. Figure 9. Age distribution of logged forest by Game Management Unit, southeastern Alaska. 55 Figure 10. Map depicting current land cover in coastal British Columbia. 58 Figure 11. Estimated percent change in mean abundance of wolves and deer under six vegetation conditions between 2015 and 2045 in Game Management Unit 2, southeastern 65 Alaska. Figure 12. Modeled estimates of percent change in mean abundance of wolves and deer based on four rates of wolf predation on deer between 2015 and 2045 in Game Management 66 Unit 2, southeastern Alaska. Figure 13. Map depicting road densities estimated by Wildlife Analysis Area and presented by Game Management Unit within the range of the Alexander Archipelago wolf in 70 southeastern Alaska. Figure 14. Map depicting road densities estimated by Wildlife Management Unit within the 73 apparent range of the Alexander Archipelago wolf in coastal British Columbia. Figure 15. Mean percent of successful hunters and trappers of Alexander Archipelago wolves 78 by type of transportation and Game Management Unit, southeastern Alaska, 1997–2014. Figure 16. Map of human settlements by population size and roads to demonstrate variation in access (e.g., road, boat) for hunters and trappers within the range of the Alexander 79 Archipelago wolf, southeastern Alaska. Figure 17. Number of Alexander Archipelago wolves harvested and reported by hunters and trappers (A) by Game Management Unit and (B) after accounting for variation in size of 82 GMU (per 1,000 km2) between 1997 and 2014, southeastern Alaska. Figure 18. Reported number of wolves harvested by regulatory year in Game Management 84 Unit 2, southeastern Alaska, 1997–2014. Figure 19. Minimum number of Alexander Archipelago wolves harvested by hunters and 85 trappers by Region between 1997 and 2012, coastal British Columbia. Figure 20. Estimated total number of wolves harvested by regulatory year in Game 87 Management Unit 2, southeastern Alaska, 1997–2014. Figure 21. Estimated percent change in mean abundance of wolves and deer under three wolf harvest guidelines (i.e., 0% reported harvest, and 20% and 30% of the estimated fall 91 population size) between 2015 and 2045 in Game Management Unit 2, southeastern Alaska. ix Species Status Assessment for the Alexander Archipelago Wolf Figure 22. Estimated percent change in mean abundance of wolves and deer under five road 93 conditions between 2015 and 2045 in Game Management Unit 2, southeastern Alaska. Figure 23. Estimated percent change in mean abundance of wolves and deer when no legal harvest of deer was permitted and when current regulations were allowed between 2015 and 94 2045 in Game Management Unit 2, southeastern Alaska. Figure 24. Mean number of reported deer harvested by (A) Game Management Unit (GMU) and (B) after accounting for variation in size of GMU (per 1,000 km2) within the range of the 95 Alexander Archipelago wolf, southeastern Alaska, 2010 and 2011. Figure 25. Estimated percent change in mean abundance of wolves and deer under three frequencies of a severe winter occurring (0.07, 0.08, and 0.10, respectively) between 2015 101 and 2045, as projected by downscaling regional climate models to Game Management Unit 2, southeastern Alaska. Figure 26. Estimated percent change in mean abundance of wolves and deer under six scenarios with variations of vegetation, wolf harvest, road density, and frequency of severe 116 winter conditions between 2015 and 2045 in Game Management Unit 2, southeastern Alaska. Figure 27. Percent change in (A) wolf abundance and (B) deer abundance across model 117 scenarios from 2014 levels, shown from the year 2000–2045, in Game Management Unit 2 x