Species Status Assessment for the Puget Oregonian (Cryptomastix Devia)

Home , Allogona, Allogona townsendiana, Ancotrema, Chondrina, Chondrina clienta, Clausiliidae

December 2019

Species Status Assessment for the Puget Oregonian (Cryptomastix devia)

Department of the Interior United States Fish and Wildlife Service Washington State Field Office 510 Desmond Drive SE, Suite 102 Lacey, Washington 98503 360-753-9440 Acknowledgements Many many thanks to all of those who took the time to introduce us to the amazing world of terrestrial gastropods, including those of you that took us to see the species, the individuals we never met that raked the litter to report their occurrence, the authors that took the time to research and describe gastropods in reports and literature, the data stewards and managers keeping everything in order, and all of you who took time to review and greatly improve this document. Without all of your wonder, knowledge, and understanding of the world around us would be diminished for the rest of us. Special thanks to: Elise Brown, Thomas Burke, Jeff Chan, Jeff Dillion James Donahey, John Fisher, Sarah Foltz Jordan, Alex Foster, Vince Harke, Paul Hendricks, Robert Huff, John Jakubowski, Edward Johannes, Tom Kogut, William Leonard, Tom McDowell, Rebecca Migala, Christine Pyle, Joanne Stellini, Kelli Van Norman

Author Contact Information An Markus Le [email protected] 360-753-7767

F. Teal Waterstrat [email protected] 360-753-7760

United States Fish and Wildlife Service Washington Fish and Wildlife Office 510 Desmond Drive SE, Suite 102 Lacey, Washington 98503 360-753-9440

EXECUTIVE SUMMARY

If we and the rest of the backboned animals were to disappear overnight, the rest of the world would get on pretty well. But if they were to disappear, the land's ecosystems would collapse …These small creatures are within a few inches of our feet, wherever we go on land - but often, they're disregarded. We would do very well to remember them. – David Attenborough, 2005 Life in the Undergrowth This document is the Species Status Assessment (SSA) for the Puget Oregonian (Crytpomastix devia) as of 2019, completed to characterize the species’ overall viability by using the three conservation biology principles of resiliency, redundancy and representation. We identify the species’ ecological requirements for survival and reproduction at the individual, population, and species levels, and describe risk factors influencing the species’ current and future condition.

The Puget Oregonian is a terrestrial snail that inhabits moist, conifer-forest habitats that include some level of deciduous tree community composition. It is strongly associated with bigleaf maple trees (Acer macrophyllum). The species is most commonly located in bigleaf maple dominant stands, or mixed maples stands growing among conifers usually Douglas-fir (Pseudotsuga menziesii), western hemlock (Tsuga heterophylla) and western red cedar (Thuja plicata), or in mixed groves of maples and other hardwoods such as black cottonwood (Populus trichocarpa) and red alder (Alnus rubra). The Puget Oregonian is most commonly found along stream and river terraces and usually absent from riparian zones prone to regular or occasional flooding. This species is often found on or under hardwood logs or other woody material, leaf litter, or under the lowest fronds of western swordfern (Polystichum munitum) that are growing near or under the maple crowns. Habitat on flat or gentle slopes are more suitable habitat than steeper slopes, perhaps because they offer more stable environments. Large diameter, older bigleaf maples provide a deep leaf litter layer and are highly suitable habitat for this species, although they may also be found under smaller diameter maples, particularly when they occur in patches or are frequently interspersed within upland conifer stands.

All Cryptomastix snails are native to the Pacific Northwest (PNW). The range of the Puget Oregonian formerly included southern British Columbia. Currently they are found in the Cascade Range and Puget Trough through the Columbia Gorge south into the foothills of the Willamette Valley. It is currently recognized as extirpated from British Columbia. The vast majority of occurrence records for this species occur in the Cispus River watershed administered by the Gifford Pinchot National Forest, Cowlitz Ranger District. Since 2014,

Miles - - several locations with known occurrences were Species Range 0 40 80 160 Habitat Suitability resurveyed for Puget Oregonian with success, however, Unsuitable no new locations or range expansions have been reported. • suitable □ State Boundaries • Species Occurrence Records Current Condition:

2 During the development of this SSA we identified 74 resiliency units. The current condition of the Puget Oregonian is characterized by 10 highly resilient units, two moderately resilient units, three units with low resiliency, and 59 units with unknown resiliency (Figure 1). Redundancy appears adequate for this species, as the known units are well distributed throughout the range.

D Sub-Basin (HUC 8) with Resiliency Units of Known Condition Miles -c:::::==--===----• D Sub-Basin (HUC 8) within Range 0 25 50 100 D State Boundaries Resiliency Unit Condition high moderate low unknown Habitat Suitability - Unsuitable - Suitable

Figure 1: Contemporary distribution of the Puget Oregonian inlcuding resiliency and represenative areas. Furthermore, the 74 resiliency units are distributed across 23 representative areas throughout the range of the species, and the 10 highly resilient resiliency units are distributed across four different representative areas, perhaps indicating adaptive capacity within the species. Nonetheless, it is important to consider the high level 3 of uncertainty surrounding the demography and distribution of this species. There is still much to learn about this species and our assessment of its current condition is based on a number of assumptions based on general Pulmonate snail biology and related PNW terrestrial gastropods. Table 1: Current resiliency condition of 15 of 74 resiliency units. Temporal Potential surrounding Number of Habitat Resiliency Resiliency Unit Name relevance of habitat connectivity occurrences Quality level* occurrence and quality High Middle Columbia-Hood-4 Contemporary Rare High Connectivity High Suitability High Middle Columbia-Hood-7 Contemporary Rare Moderate Connectivity Moderate Suitability High Puget Sound-1 Contemporary Rare High Connectivity High Suitability Puget Sound-6 Contemporary Rare Not Suitable Isolated Low High Snoqualmie-1 Contemporary Rare Moderate Connectivity Moderate Suitability Snoqualmie-2 Mid-Century Rare Suitable Moderate Connectivity Low High Upper Cowlitz-2 Contemporary Common Moderate Connectivity High Suitability High Upper Cowlitz-3 Contemporary Common High Connectivity High Suitability Low Upper Cowlitz-5 Contemporary Rare Low Connectivity Low Suitability High Upper Cowlitz-6 Contemporary Common High Connectivity High Suitability High Upper Cowlitz-7 Contemporary Common High Connectivity High Suitability High Upper Cowlitz-11 Contemporary Rare High Connectivity High Suitability High Upper Cowlitz-13 Contemporary Rare High Connectivity High Suitability High Upper Cowlitz-15 Contemporary Rare High Connectivity High Suitability High Upper Cowlitz-16 Contemporary Rare High Connectivity High Suitability *There are 59 additional identified and mapped resiliency units for this species which the Service was unable to determine a current resiliency condition at this time due to the limited available information (Appendix III).

Future Condition The SSA process for the Puget Oregonian was limited in its ability to assess the current viability of the species across its range. The Upper Cowlitz sub-basin contains the majority of observational and biological information on the species, making it the basis for most of our assumptions about the species’ viability. Based upon available survey data from the Upper Cowlitz sub-basin, we understand the species to be fairly common and well distributed within suitable habitat within the sub-basin. Outside of this sub-basin, species records are relatively rare, limiting our ability to more comprehensively analyze the resiliency, redundancy, and representation of the species across its range. However, we were generally able to identify habitat characteristics known to be important to the species. These habitat characteristics include bigleaf maple stands or mixed bigleaf maple stands and microhabitat features such as large woody debris, swordferns, and leaf litter that contribute to moist and cool conditions on the forest floor. To assess future conditions of the species out to the middle of the century, we considered the effects of climate change on the species’ modeled habitat requirements. We also considered other relevant risk factors, including

4 human population growth in the PNW, forest management, and disease. We identified four future scenarios, one which anticipates no change from current condition, and three scenarios which anticipate various levels of decline in habitat conditions. Similar to the assessment of current condition, we did not have enough information to assess the future condition of the additional 59 resiliency units. There is a wide range of possibilities for the future of the 15 resiliency units for which we did have sufficient information. For example, at this time we are unable to assess the potential severity and extent of maple die-off disease in the future. We expect the resources of the species to diminish in quantity and quality as stressors increase in the future. Overall, we expect the viability of the species to decline under the future scenarios, but at the middle of the century we expect the species will still persist. Summary Although the SSA provides a useful framework by which species viability can be assessed, the quantity and quality of information and data about the Puget Oregonian made it problematic to conduct a rigorous evaluation of Resiliency, Representation, and Redundancy and difficult to provide a precise prediction of present and future conditions for the species. However, we know there were more habitat and locations where the snail occurred in the past compared to present times, and we know that predicted future conditions will present challenges for the species and the ecosystem for which it depends.

5 Table of Contents

EXECUTIVE SUMMARY ...... ii INTRODUCTION ...... 1 Background ...... 1 Analytic Framework ...... 1 Status Determinations by other agencies/organizations ...... 2 SPECIES INFORMATION ...... 3 Taxonomy: ...... 3 Species Cryptomastix devia ...... 4 Range ...... 6 Species Distribution Model and Current Range ...... 9 Modeling Variables Considered ...... 9 Modeling of Puget Oregonian Habitat ...... 10 Current Range of the Puget Oregonian ...... 12 HABITAT ...... 13 General Habitat ...... 13 Habitat Components ...... 15 LIFE HISTORY OF THE PUGET OREGONIAN...... 17 Longevity of the Puget Oregonian ...... 17 Reproductive and Developmental Life History ...... 19 The Active State of the Puget Oregonian ...... 21 The Roosting and Dormant State of the Puget Oregonian...... 21 LIFE ON THE FOREST FLOOR: ...... 22 Movement and dispersal: ...... 22 Diet and foraging: ...... 23 Respiration: ...... 24 SUMMARY OF INDIVIDUAL, POPULATION, AND SPECIES NEEDS ...... 24 Individual needs: ...... 24 Population needs ...... 25 Species needs ...... 26 FACTORS INFLUENCING THE SPECIES ...... 26 Stressors to the species needs: ...... 27 Forest management ...... 27 Resiliency ...... 32 Representation ...... 34 Summary of Current Condition ...... 36 FUTURE CONDITIONS ...... 38 6 Analysis of Future Conditions ...... 39 Risk Factors ...... 40 Future Scenarios ...... 42 Assumptions of Future Conditions ...... 43 Summary of Future Condition ...... 47 Synthesis ...... 49 Literature Cited ...... 50 Personal Communications ...... 56 Appendix I Molluscan Taxonomy, Physiology, and Ecology ...... 1 Appendix II: Maxent model for Puget Oregonian ...... 1 Analysis of omission/commission ...... 1 Pictures of the model ...... 3 Response curves ...... 4 Analysis of variable contributions ...... 5 Raw data outputs and control parameters ...... 7 Appendix III – Resiliency Unit Current and Future Condition Scoring ...... 1 Appendix IV Climate Change ...... 1

List of Figures: Figure 1: The above figure depicts how the Service uses Species Status Assessments to inform actions under the ESA. ……………………………………………………………………………………………………………1 Figure 2: Images of Cryptomastix devia, the Puget Oregonian, in the Cispus Watershed of Gifford Pinchot National Forest, Lewis County, WA on May 10, 2019…………………………………………………………5 Figure 3: Terrestrial Mollusk Providences of North America…………………………………………………..6 Figure 4: Published range of the Puget Oregonian……………………...………………………………………6 Figure 5: Maxent Modeling output……………………………………………………………………………..11 Figure 6: Current range of the Puget Oregonian in Washington and Oregon……….....……………………….13 Figure 7: Stand level habitat of the Puget Oregonian…………………………………………………………..14 Figure 8: A sampling of habitat where Puget Oregonian were located in May 2019…………………………..15 Figure 9: Basic life cycle of the Puget Oregonian ………………………..……………………………….……19 Figure 10: Known resource needs of Puget Oregonian individuals and populations…………………………..26 Figure 11: Graphic depiction of Puget Oregonian resiliency units……………….…………………………….37

List of Tables: Table 1. Puget Oregonian occurrence records in relation to elevation…………………………………………9 Table 2. Puget Oregonian occurrence records in relation to slope……………………………………………..9 Table 3. Puget Oregonian occurrence records in relation to normal annual precipitation……………………10 Table 4. Puget Oregonian occurrence records in relation to the bigleaf maple range…………………...... 10 Table 5: Individual Resource Needs of the Puget Oregonian.………………………………....……………....25

7 Table 6: Sub-basin (HUC 8) representative areas for the Puget Oregonian in Washington and Oregon……….31 Table 7: Categorical definitions used to assess the current condition of the resiliency units of Puget Oregonian ……………….………………………………………………………………………………………………….32 Table 8: Known Current Condition of the resiliency units of the Puget Oregonian …………………………...34 Table 9: Representative areas for the Puget Oregonian ....……………………………………………………...34 Table 10: The four future scenarios used to estimate future conditions at each site…………………………....42 Table 11: Conditions of 15 known Puget Oregonian resiliency units currently and under future scenarios…...44 Table 12: Summary of current and future condition of Puget Oregonian in future scenarios……………….…48

8 INTRODUCTION

Background

In 2008, the U.S. Fish and Wildlife Service (Service) was petitioned to list the Puget Oregonian (Cryptomastix devia) as threatened under the Endangered Species Act, as amended, (ESA). In 2011, the Service issued a positive 90-day finding on the petition to list the species, having determined that the petition presented substantial scientific or commercial information indicating that listing the Puget Oregonian may be warranted. The 90-day finding noted the following potential threats: the present or threatened destruction, modification or curtailment of its habitat or range resulting from forest management and land conversion to agriculture; and other natural or manmade factors as affecting its current existence resulting from high intensity fire, and from increased tree mortality due to various causes associated with climate change. This document presents the species status assessment (SSA) for the Puget Oregonian completed to characterize the species’ overall viability as of 2019 by using the three conservation biology principles of resiliency, redundancy and representation. We identify the species’ ecological requirements for survival and reproduction at the individual, population, and species levels, and describe risk factors influencing the species’ current and future condition. Analytic Framework

Using the SSA (Figure 1), the Service can characterize the status of the species in terms of its resiliency, redundancy, and representation (Wolf et al. 2015, entire) and use the document as the foundation for numerous aspects implementing sections of the ESAESA. We define Resiliency, Redundancy and Representation below.

• Not wa rr.mted

Llsl Ing dedi lOfl

c...-.11 n... tl<,,, Jrttte,u•.oc6Kf..~e ln 1-i~J.Cn and OQwnlistedor 1;q:-u.e~tili;n .:lfQf'b. d@list@d

SSA Soe-«ol"f'I • Curren t. ccir.;fo1lon Fub;l!II! tonrlllietl

Ad<1pteo 10 tM declsron conte-~t ~ocl 11pd.-it d to in( lude new data and information Figure 1: The above figure depicts how the Service uses Species Status Assessments to inform actions under the ESA.

Resiliency describes the ability of populations to withstand stochastic events (arising from random factors). We can measure resiliency based on metrics of population health; for example, birth versus death rates and population size. Highly resilient populations are better able to withstand disturbances such as random 1 fluctuations in birth rates (demographic stochasticity), variations in rainfall (environmental stochasticity), or the effects of anthropogenic activities. Representation describes the ability of a species to adapt to changing environmental conditions. Representation can be measured by the breadth of genetic or environmental diversity within and among populations and gauges the probability that a species is capable of adapting to environmental changes. The more representation, or diversity, a species has, the more it is capable of adapting to changes (natural or human caused) in its environment. In the absence of species-specific genetic and ecological diversity information, we evaluate representation based on the extent and variability of habitat characteristics across the geographical range. Redundancy describes the ability of a species to withstand catastrophic events. Measured by the number of populations, their resiliency, and their distribution (and connectivity), redundancy gauges the probability that the species has a margin of safety to withstand or can bounce back from catastrophic events (such as a rare destructive natural event or episode involving many populations). Because there are multiple (redundant) sites within the sub-basins or grouping of sub-basins we will describe intra sub-basin or assessment unit redundancy for each geographic area. The Service has grouped Puget Oregonian into sub-basins to evaluate the overall viability of the species. These sub-basins or resiliency units are how we assess representation. Redundancy across these units is described at the range-wide scale. Status Determinations by other agencies/organizations

Other entities beyond the Service are concerned with the conservation of rare species. In this case, this includes the U.S. Departments of Agriculture and Interior, Washington Department of Fish and Wildlife (WDFW), Oregon Biodiversity Information Center (ORBIC), and various non-profit conservation organizations including NatureServe and the Xerces Society for Invertebrate Conservation. Globally the Puget Oregonian is ranked G2, “Imperiled” by NatureServe standards (ORBIC 2019, p. 6). The U.S. Department of Agriculture considers the Puget Oregonian a Sensitive Species in Region 6 of the U.S. Forest Service (USFS) as does the Department of Interior Bureau of Land Management (BLM) (Foltz Jordan and Black 2015 p. 6). For the Forest Service in Oregon and Washington, the species is also considered to be a “Survey and Manage” species under the Northwest Forest Plan (1994), and accompanying addendums (2001). The species is currently “Category A” under Survey and Manage, which requires the USFS to conduct surveys for the species whenever habitat- disturbing activities are proposed, and to manage (protect) sites discovered from those or other surveys. There are some exemptions to this Survey and Manage requirement (Pechman 2006) notably thinning in stands less than 80 years of age. In Oregon, it is considered critically imperiled (ORBIC 2019, p. 6), and in Washington State it is not considered species of greatest conservation concern (WDFW 2014). It is considered extirpated from Canada by Committee on the Status of Endangered Wildlife in Canada (COSEWIC 2013, pp. iv – vi).

NatureServe is a non-profit that ranks organisms in the United States with State Ranks (S#) which describe the rarity of a species within each state's boundary. These Ranks begin with the letter representing the scale they are assessed. Global (G#), National (N#), and State (S#) ranks all use a 1to5 and x ranking system, 1 being the most imperiled, 5 being secure, and “x” representing the organism is extirpated (Master et al. 2012, pp. 48-50). Global and State ranks for the Puget Oregonian are summarized below. Although these rankings are broadly accepted they are not equivalent or comparable to the evaluation of a species as threatened or endangered under the ESA. Globally, the Puget Oregonian is ranked G2, Imperiled (ORBIC 2019, p. 6) defined as at a high risk of extinction due to very restricted range, very fewI populations (often 20 or fewer), steep declines, or other factors. In Oregon, it is a State Rank 1 species (ORBIC 2019, p. 6). State Rank 1 species are considered critically imperiled because of extreme rarity or because it is somehow especially vulnerable to extinction or extirpation,

2 typically with five or fewer occurrences. In Washington State, Nature Serve considers the specie’s State Rank 2 or 3. A state Rank of 2 is defined as “Imperiled” because of rarity or because other factors demonstrably make it very vulnerable to extinction (extirpation), typically with 6-20 occurrences. A state Rank of 3 indicates it is “rare, uncommon or threatened,” but not immediately imperiled, typically with 21-100 occurrences. WDFW does not consider it to be a species of greatest conservation need (SGCN) because it is widely distributed and locally common in the Cowlitz Valley (WDFW 2014). If it became a SGCN the State would collect occurrence data for periodic review of its status but provide no protections. The Canadian Province of British Columbia considers the species extirpated from the Province and Canada based on the few (3) historic records and negative detections from contemporary resurvey and inventory efforts in suitable habitat (COSEWIC 2013, p iv – vi). This informed and concurs with the NatureServe rank of SX (extirpated) in British Columbia (http://explorer.natureserve.org/servlet/NatureServe?searchName=Cryptomastix+devia, accessed May 16, 2019).

SPECIES INFORMATION Taxonomy:

Cryptomastix devia - Puget Oregonian (Gould 1846) Etymology as described in Forsyth 2004 (p. 152) Cryptomastix: Hidden flagellum (to anatomically distinguish it from the eastern and central North American genus Triodopsis.) devia: out-of-the-way or solitary

Puget Oregonian (Cryptomastix devia) snail is considered a valid species (ITIS, May 15, 2019, Turgeon et al, 1998 p. 150). It formerly, and still commonly, is placed taxonomically as follows (with some reclassification of the Pulmonate grouping): Phylum: Mollusca Class: Gastropoda Subclass: Pulmonata Order: Stylommatophora Family Polygridae Genus: Cryptomastix (Pilsbry 1939, p xvii) Species: devia (Gould 1846) (ITIS, May 15, 2019, Turgeon et al. 1998, pp. 137-150)

The above “classic” Linnaean taxonomic structure reflects what is published in the authoritative taxonomic references for this species (ITIS, May 15, 2019, Turgeon et al, 1998 p. 150). However a large restructuring of higher classifications and clades of gastropods have occurred since that time and is presented below (Bouchet and Rocroi 2005, entire; Poppe and Tagaro 2006, entire; Bouchet et al. 2017, entire). In the revised and accepted phylogenetics of gastropods the Puget Oregonian belongs to:

Phylum: Mollusca Class Gastropoda Clade Heterobranchia Informal Group Pulmonata Clade Eupulmonata Clade Stylmmatophora Informal Group Sigmurethra 3 Superfamily Helicoidea Family Polygridae (Bouchet and Rocroi 2005, p. 264-270, 410; Bouchet et al. 2017, p. 367).

It is important to understand the current phylogenetic system of this species. Fortunately most general biological information presented in this document will be discussed at the level of the Pulmonata, which is still a valid classification in both classic and current taxonomic structures. Thus it will be important to understand the current phylogenetics of gastropods to reflect the best available scientific information and the “classical” structure in literature to understand the physical and biological needs of the species. A detailed account of Mollusca taxonomy, physiology, and ecology relevant to the Puget Oregonian can be found in Appendix I.

Species Cryptomastix devia

Taxonomic history This species was first described by Gould as Helix devia in 1846 (Gould 1846, p.165). Burke (1999, p. 3) further lists 6 revisions of genus name between 1846 and 1940. The species designation, devia, has been valid, with one early exception, throughout those transitions (Burke 1999, p. 3). Physical description The Puget Oregonian is significantly larger than any others in the genus, and is readily distinguished from other Cryptomastix by a combination of mature shell size (20 to 24 mm wide by 12.5 to 16 mm high, with 5.5-6.0 whorls) and its distribution (Burke 2013, pp. 157-158). In addition, a prominent white parietal tooth is typically observed on median half of the basal lip margin (Figure 1b). The umbilicus is partly covered. The shell spire is elevated; the periphery is rounded. The color of the shell is medium to dark brown; the peristome is white to tan with irregular, shallow spiral striae which appear as wavy lines under the magnification (Figure 1a- e). The depressed globose- shaped shell has a heavy, broadly reflected apertural lip which partially or mostly covers the umbilicus (Figure 1a – d). The narrow umbilicus is one- eighth to one-tenth the diameter of the shell. Cryptomastix are generally hairless as adults. Immature Puget Oregonians have short, hooked microscopic bristles on the dorsal surface which are lost by adulthood and are soon lost on preserved shells. Juvenile Puget Oregonians also lack the reflected apertural lip (Figure 1e).

Figure 2: Images of Cryptomastix devia, the Puget Oregonian, in the Cispus Watershed of Gifford Pinchot National Forest, Lewis County, WA on May 10, 2019. A. Active adult C. devia. B. Recurved lip aperture (left), single prominent white parietal or apertural tooth (right) and overall medium to dark brown shell color are identifying characteristic of a mature individual. C and D. Size of mature snail shell width ~20 to 24 mm adult hand for scale. Faint mucus trail can be seen on hand behind moving snail. E. Live juvenile Puget Oregonian lacking recurved lip and parietal tooth. (Images USFWS)

SPECIES SIMILAR IN APPEARANCE IN THE RANGE OF THE PUGET OREGONIAN Burke (1999, pp. 3-4; 2013, p. 158) identifies similar appearing species that may be confused with the Puget Oregonian and physical characteristics that can be used to distinguish them in the field. Most notable are the large size and presence of apertural tooth (Figure 2b). Below are the similar species Burke lists along with a brief distinguishing statement. • Cryptomastix hendersoni: slightly smaller than Puget Oregonian and typically lack apertural tooth. • Cryptomastix germana: smallest of the Cryptomastix reaching about 8 mm in size and retains curved bristles on shell as an adult • Allogona ptycophora: similar in size and form, but lacks parietal tooth • Allogona townsendiana: larger in size and similar in form, but lacks parietal tooth • Monadenia fidelis (juvenile): juveniles of this species can be confused with juvenile Puget Oregonian, but shell coloration, patterns, and sculpturing are typically distinct. .

Genetics: Its taxonomic status is considered valid based on genetic and morphological characteristics (ITIS, May 15, 2019;; Turgeon et al. 1998, p. 150; Perez 2014, 17). There is likely some genetic variation within the species as it has a large range from the coast range of Oregon to (formerly) the Fraser Valley and extreme southern portion of Vancouver Island, British Columbia (see range section below). To the best of our understanding, no genetic evaluation within the species has been conducted. The species is apparently restricted by individual dispersal 5 ability, and as a species was probably subject to many geologic events that isolated populations geographically in glacial refugia south of the current Skagit River Valley and between the Puget lobe the Cordillearan Ice Sheet and glaciers in the Cascades during the latest glacial period (Bretz 1913; Easterbrook 1967, p. 13)

Range Terrestrial snails are broadly distributed worldwide, however, in the northern hemisphere post-glacial period Pulmonate snail colonization has been limited (Barker 2001, p. 84; Solem 19841984, pp. 8-9). In North America, land snail biogeography was divided into 10 provinces by Pilsbry (1948, p.xli, xlvi), which are still used to delineate regional boundaries in contemporary work in the western United States (Frest and Johannes, 1993 pp. 6-7; Burke 2013, pp. 17-19) (Figure 3). Species of the genus Cryptomastix are restricted the Pacific Northwest (PNW) and in the Oregonian and Washingtonian mollusk provinces of the northwestern United States (Burke 2013, p. 156-169). The identified

~;t(';JF'IC. (I) range of the Puget Oregonian to date are based on OCl!A!II (0) (W) observed and confirmed specimens of the species dating I , ' \ back to the early 1900s (Ovaska and Forsyth 2002, p. 6; on&iON I I F-- - ...... WYOMUIG Burke 2013, p. 159; Foltz Jordan and Hoffman Black I I (R) 2015, pp. 9-11, 31), and appears to occur almost entirely ...-'°'C'AUl'OltJlll A Nl!VADA within the Oregonian province which extends from coastal \ I {C) \ \ British Columbia west from the Cascade Mountain Divide \ just into extreme northern California, with a very small Con l:. inental Divide" portion of its range that extends into the Washingtonian Figure 3: Terrestrial Mollusk Providences of North America: (C) Californian; (I) Interior: (O) Oregonian; providence in the Columbia River Gorge (Figure 4). (R) Rocky Mountain; (W) Washingtonian. Defined by To help understand the current potential range of the Pilsbry (1948) and copied from Burke 2013. species, the current distribution of the species within its range, and to inform the SSA, the Service conducted a British Columbia distribution mapping exercise. Survey and Occurrence Information Currently, the Service is aware of 228 locations where the Puget Oregonian has been recorded. West to east, all but three of these records can be found from the eastern Olympic Peninsula to the western slopes of the Cascade Mountain Washington Range in Washington State and within Oregon’s Coast Range and Willamette Valley. Two of the remaining records on the eastern slopes of the Cascades are located within the Okanagan-Wenatchee National Forest, while the third is located within the Columbia River Valley (Jordan Foltz and Hoffman Black 2015, pp. 9-11). Species occurrences have been as far north as the Mount Baker-Snoqualmie National Oregon N Forest near Darrington, Washington, and as far south as the 25 50 ,oo Mount Hood National Forest. Miles Figure 4: Published range of the Puget Oregonian coarsely adapted from Ovaska and Forsyth 2002, p. 6 and Burke 2013 p. 158. In the 1990’s documented Puget Oregonian sites were scattered in eastern central Puget Sound (Frest and Johannes 1995, p 229). Overall populations (number of sites or number of individuals) were considered to be

6 on downward trend (Frest and Johannes 1995, p 229). Although there has been an increase in the number of individuals observed and locations where the species has been found since the 1990’s, there has been no subsequent demographic summary (Flotz-Jordan and Hoffman Black 2015, p. 11). The largest number of Puget Oregonians found at any one site prior to 1995 was six (Burke 1999. p. 7). To the best of our understanding the greatest concentration to date is eleven individuals at a single site (NRIS database, April 18, 2019). Throughout most of the species’ range, available species occurrence records are diffuse, and include some observations in urban locations. The biggest exception is the Upper Cowlitz sub-basin, which contains a cluster of 164 species occurrence records (Jordan Foltz and Hoffman Black 2015, pp. 9-11, Xerces dataset). The three historic records for the Puget Oregonian in Canada, all observed before 1905; two recorded on Vancouver Island, and one recorded on mainland British Columbia. Only the occurrence recorded on Vancouver Island is considered to be reliable, while the other two records are of questionable validity (Ovaska and Forsyth 2002, pp. 5-8). Recent searches for the species, occurring between 1986 and 2006, have not found the species on Vancouver Island or mainland British Columbia. As a result, the species is considered extirpated from Canada since 2002 (Ovaska and Forsyth 2002, pp. 5-8). Of the available species records, 193 out of 228 are located on federal lands, 182 of which are located on USFS lands. The remaining records on federal lands are spread between the BLM, and the Service. Of the remaining data, four records are located on state lands, and 31 are located within other land ownerships, including county, municipal, and private lands. Most of the available species records, 200 out of 228 were observed after the adoption of the Northwest Forest Plan (NWFP) in 1994, of which 185 are located on federal lands. Data from before the adoption of the NWFP date as far back as 1906. Additionally, 15 of the known observations occurred at an unknown date, and are located mostly on local and private landownership (Jordan Foltz and Hoffman Black 2015, p. 31). Very few, if any, of these records have been resurveyed over time, so information on the persistence of the Puget Oregonian on the landscape is limited. There are several locations where the species has been confirmed to still persist is the upper Cowlitz sub-basin. These include sites near the Cispus Learning Center, where Service biologists observed the species in 2019 during a field visit, 20 years after the species was reported in that location (Le and Waterstrat 2019, pers. obs.) and a site in Packwood where it has been observed over may years (Kogut, T. pers. comm. 2019). The location and timing of most species records within federal lands after 1994 is due to the adoption of the NWFP in 1994, which designated the Puget Oregonian as a “Survey and Manage” species. Due to its status, the USFS must survey for the species when a habitat-disturbing activity is proposed in an area. The BLM was also under the Survey and Manage standards until 2006, when the BLM changed its Resource Management Plans to no longer include this requirement. Many of the extant species occurrence data have been recorded as a result of these surveys conducted by the USFS and BLM. The NWFP Survey and Manage Protocol requires the surveys be conducted to habitat that may be disturbed by a proposed project. The survey protocol requires small project areas (1/4-1 acres) to be surveyed a minimum of 20 minutes, projects that are one to 10 acres a minimum of 45 minutes, and projects disturbing more than10 acres, at least for one hour per 10 acres. In these larger project areas, for every 10 acres, two smaller areas of highly suitable habitat are surveyed for a minimum of 20 minutes each, while the remaining 20 minutes is spent meandering throughout the remainder of the 10 acres, spot checking higher-quality habitat found along the way. (Duncan et al. 2003). Two surveys, one recommended in the fall rainy season, or later spring, are required during appropriate environmental conditions, and should be at least 3 weeks apart. This protocol is not designed to identify every occupied area within an action footprint, and may result in the species not being found in locations where it is present. Additionally, surveys following this protocol are only required where land management is to occur. As a result, surveys for the species are limited on unmanaged areas on federal lands. Outside federal lands, observations of the Puget Oregonian have been recorded mostly by malacologists conducting individual inventories to further knowledge of the species. Thus, although the survey information reported from Federal agencies can confirm presence of

7 the Puget Oregonian, it is inherently limited in scope and cannot be used as a basis for reasonable abundance or distribution estimates for the species. A significant portion of occurrence records come from the Natural Resource Manager (NRM), formerly called the Natural Resource Information System (NRIS), and the Geographic Biotic Observations (GeoBOB) databases, which contain survey data on survey and manage species, including the Puget Oregonian, within USFS and BLM lands, respectively. Until 2006, when the BLM revised its Natural Resource Management Plans, the GeoBOB database served as the data repository for all survey and manage species data, and thus also contains a significant amount of information recorded on USFS lands. As the data repository for BLM Survey and Manage species data in Washington and Oregon, the database also contains locations where surveys on federal lands have occurred in the two states. In Oregon, the database indicates that surveys for terrestrial mollusks have occurred on BLM lands in the northwestern portion of the state, with only a few occurring south of the Siuslaw National Forest, and none occurring east of the Cascade Mountain Range. In Washington, very few surveys have been recorded in the database, including some locations within the San Juan Islands National Monument, and some on the Umatilla National Forest in the far southeastern portion of the state. Other than the few surveys documented in the Umatilla National Forest, no survey efforts were documented within USFS lands in the GeoBOB database (BLM 2019). However, due to the presence of occurrence records within National Forests in both Oregon and Washington, even where survey effort data are lacking, it is certain that survey efforts far exceed the amount captured in the GeoBOB database. While survey efforts under the NWFP are likely available in non-digital formats at the USFS, and BLM those data are not easily accessible to the Service at this time. Considering the brief terrestrial mollusk protocol survey, the concentration of survey data on managed lands, and the lack of available information on where surveys have occurred, at this time it is impossible to conclude that a lack of Puget Oregonian occurrence data indicates the absence of the species. Although the Puget Oregonian is a relatively large snail, it is easily overlooked during surveys due to its cryptic brown coloration, which blends in with its leaf litter surroundings (Jordan Foltz and Hoffman Black 2015, p. 11). However, because of the species’ relatively distinct physical characteristics and habitat preferences, it is relatively easy to identify when found. The species’ low detectability, limited number of surveys, and the limited scope of surveys hinder our ability to estimate abundance, density, or persistence of the species throughout most of its range, though the abundance and density of the species are likely low. The exception is the Upper Cowlitz sub-basin, where it can be roughly inferred that the species is relatively abundant, and has persisted throughout time, due to the relatively high number of species records present. Due to the lack of species data collected at consistent locations across time, it is impossible to determine the demographic characteristics of the species, such as population trends, and reproductive rates. In addition to the 228 Puget Oregonian data records collected from federal surveys and researchers, some species records are available from publicly accessible databases, including iDigBio, BISON, iNaturalist, and the Global Biodiversity Information Facility. These databases contain an additional 35 species records, most of which are located within the same geographic extent as the data described above. These records were collected between 1870 and 2018. Six records of the species from public databases are located much further east than expected, including one in eastern Washington, four in Idaho, and one in West Virginia (iDigBio 2019, USGS 2019, iNaturalist 2019, GBIF 2019). Due to the incorporation of citizen science data into these databases, the relatively dry environments found in those locations, and the lack of any other validated occurrence data in those locations, the validity of those occurrence records is questionable and likely represent other species in the Polygridae family (i.e.: Cryptomastix or Allogona spp) misidentified as Puget Oregonian. The species records obtained from these publicly accessible databases are not included in the analysis of the species range below, as these 35 records were not available to the Service at the time the analysis was conducted. However, all but the six questionable data records fall within the modeled range and habitat descriptions described below, corroborating the results of the range analysis.

8 Species Distribution Model and Current Range

The Service examined occurrence data for Puget Oregonian (Xerces Society 2019, BLM 2019) in relation to several environmental variables, including elevation, slope, precipitation, and the range of bigleaf maple which the species is closely associated with.

The adoption of the NWFP in 1994 by several federal land management agencies, including the USFS, and the BLM, has had wide-ranging implications on the management and conservation of wildlife within the range of the northern spotted owl (Stix occidentalis caurina), including a suite of terrestrial mollusks such as the Puget Oregonian. This included the development and adoption of standardized protocols for identification and occupancy of terrestrial mollusks (USDA and DOI 1994, pp. 58, C-59 C-60). Because of the increased effort and standardized protocols and reporting, the Service considers points observed before 1994 as valid historic records and points after 1994 as new current records for the purposes of this SSA. For the purposes of habitat analysis and modeling, we did not exclude historic data, as it is unlikely that conditions have worsened for the species on federal lands due to the implementation of the NWFP. While some historic occurrence records outside federal lands are now located in urbanized areas, we did not exclude those locations from the range analysis either, as the snail may be able to persist in such landscapes in small intact patches of habitat within areas such as parks and greenbelts.

Modeling Variables Considered

Multiple sources have noted that the species tends to be found below certain elevations within suitable habitats, though stated upper elevation limits vary between 2,000 feet (610 meters) and 2,700 feet (823 meters) above sea level (Burke 2013, p. 158; Jordan Foltz and Hoffman Black 2015, p. 10: USDA, 2008, p. C-39). We found that 221 out of 228 available occurrence records did occur below 2,700 feet (823 meters) (Table 1).

Table 1. Puget Oregonian occurrence records in relation to elevation. Upper elevation Upper elevation Number of Percentage of (feet) (meters) observations observations 2700 823 221/228 96.9% 2500 762 215/228 94.3% 2000 610 193/228 84.6%

In relation to slope, we found that all recorded occurrences occurred on slopes below 40 degrees, with the vast majority occurring on slopes below 30 degrees (Table 2). The occurrence of the species mostly on more gradual slopes may indicate a preference for sites that are more geomorphically stable and can retain moisture for longer periods of time. However, this trend may also be present due to survey bias, as steep slopes can be difficult to survey.

Table 2. Puget Oregonian occurrence records in relation to slope. Slope (degrees) Number of Percentage of observations observations <10 108/228 47% <20 176/228 77% <30 220/228 96% <40 228/228 100%

Most of the occurrence records are in locations where normal precipitation totals at least 40 inches (in) (101.6 centimeters (cm)) per year. All records but one occur in locations with normal annual precipitation exceeding 35 in (89 cm), with the remaining record (located in the Columbia River Valley) receiving about 16 in (40.6 cm) of annual precipitation (Table 3). The overall high amount of normal annual precipitation at the majority of the species occurrence records is indicative of the snail’s need for moisture to survive.

Table 3. Puget Oregonian occurrence records in relation to normal annual precipitation. Annual Number of Percentage of Precipitation observations observations (inches) ≥40 216/228 94.7% ≥37 224/228 98.2% ≥35 227/228 99.6%

The snail is thought to be strongly associated with bigleaf maples, as the tree provides leaf litter and shelter on the ground for the snail (Kelley et al. 1999, p. 5; Burke 1999, pp. 1, 5-6; Foltz Jordan and Black 2015 pp. 4, 9, 11-12). In general, the species occurrence data are located mostly within the range of the bigleaf maple. However, the range of the bigleaf maple appears to be much larger than that of the snail, extending from southern British Columbia to California. We found that 215 out of 228 available occurrence records were located within the range of the bigleaf maple, as described by Little (1971, Map 95-N). The remaining occurrence records occur within relatively short distances of the described bigleaf maple range, with 226 out of 228 records located within about 4 miles (7 km) of the boundary. Due to the generalized, coarse nature of the boundary for the bigleaf maple range, it is still possible that bigleaf maple may be found in low quantities at these sites. The remaining two points, located about 24 miles (39 km) and 30 miles (48 km) from the boundary of the bigleaf maple range, are also among the few occurrence points occurring on the eastern side of the Cascade Mountain Range (Table 4).

Table 4. Puget Oregonian occurrence records in relation to the bigleaf maple range. Distance to Number of Percentage of Bigleaf maple observations observations range (km) 0 215/228 94.3% <1 217/228 95.2% <3 224/228 98.2% <7 226/228 99.1% <50 228/228 100%

Modeling of Puget Oregonian Habitat

10 We used a Maxent model (Phillips et al. 2017, entire) to estimate the habitat characteristics and current range of suitable habitat for the Puget Oregonian within the entirety of Washington and Oregon. The input environmental variables for the model were aimed to capture habitat characteristics that are likely important to the survival of the species, including temperature, moisture, amounts of vegetative cover, and types of vegetative cover. Due to lack of geographic data directly related to this information, most of the variables used were proxy variables, as described below. While the general habitat characteristics and thresholds described above are useful in understanding the habitat needs of the species, we did not impose upper limits for elevation and slope, lower limits for precipitation, or the range of the bigleaf maple within the model, in order to allow the model to create a habitat map unbiased by our expectations for suitable habitat, as described above, and identify potentially suitable habitat where surveys for the Maxent Model Output Miles - - 0 40 80 160 species have not yet occurred or where the species are Habitat Suitability not present due to historic geologic events (Figure 5). 0.87 For detailed model results, refer to Appendix II. 0 □ State Boundaries Elevation: The elevation data used in the model came • Species Occurrence Records from the NASA Shuttle Radar Topography Mission (SRTM) (USGS 2006). The SRTM dataset provides global elevation data at a spatial resolution of 1 arc-second (98.5 ft or 30 meters). Model results indicate that occurrence of the species is most likely at lower elevations, and unlikely to occur at elevations above about 3,281 ft or 1000 meters, corroborating the observations described above. Elevation was among the most important variables in the Maxent model, as indicated by jackknifing results and contribution to model gain.

Slope: Slope data were derived from the SRTM elevation dataset, and thus have a spatial resolution of 98.5 ft or 30 meters. Model results indicate that the probability of species presence is generally highest on flat slopes, with probabilities declining as slopes get steeper. These results also corroborate observations described above. However, this variable was among the two least important variables in the model, as indicated by jackknifing results and contribution to model gain. Furthermore, due to the spatial scale of 98.5 ft (30 meters), the presence of small flat areas, such as terraces and benches on hill sides that may support the species, would not be reflected in this dataset.

Topographic Position Index: Topographic Position Index (TPI) data were derived from the SRTM elevation dataset, and thus share its spatial resolution of 98.5 ft or 30 meters. As implied by its name, TPI indicates the position of a location along a slope by comparing the elevation of the location to the average elevation of its surroundings. Positive TPI values usually indicate ridge or hill tops, negative values indicate valley bottoms, and values near zero indicate hillsides or flat areas. The TPI of a location is dependent on the scale at which it is calculated (Gallant and Wilson 2000). For the purposes of this model, a TPI dataset was calculated at a relatively fine scale (annulus of 328 ft (100) meters to 656 ft (200 meters)) to capture localized topography, and one was calculated at a coarser scale (annulus of 1.0 km to 1.5 km) to capture larger landforms. Model results show that the species is most likely to occur at negative coarse-scale TPI values, indicating species preference for locations in or near valley bottoms. While not as important as elevation, precipitation, or Normalized Difference Vegetation Index (NDVI), this variable was important in the Maxent model, as indicated by jackknifing results and contribution to model gain. However, TPI calculated at the finer scale exhibited no strong trends, and did not contribute significantly to the model.

11 Normalized Difference Vegetation Index: Due to the tendency of plants to absorb solar radiation in the photosynthetically active radiation spectrum and reflect radiation in the near-infrared spectrum, multispectral satellite imagery can be leveraged to calculate the NDVI. High values of this index usually indicate locations that are densely vegetated (Rouse et al. 1974). The NDVI dataset used for this model was derived by the USGS from multispectral data obtained from the Moderate Resolution Imaging Spectroradiometer (MODIS) satellite in August 2018 (USGS 2018). The 820 ft (250 meter) resolution dataset was downscaled to 98.5 ft or 30-meter resolution to match the spatial resolution of the topographic datasets, using bilinear interpolation. Model results show that probability of species presence rises as the NDVI increases, signifying that the species tends to prefer densely vegetated areas, such as forests. NDVI was among the most important variables in the Maxent model, as indicated by jackknifing results and contribution to model gain.

Precipitation: Normal annual precipitation data were obtained from the Parameter-elevation Regressions on Independent Slopes Model (PRISM) dataset (PRISM Climate Group 2015). The dataset shows 30-year normal values, as calculated from 1981 to 2010, and has a spatial resolution of 262.5 ft (80 meters). For the purpose of this model, the dataset was downscaled to 98.5 ft (30-meter) resolution to match the spatial resolution of the topographic datasets, using bilinear interpolation. Model results show that the species is most likely to be present in areas where annual precipitation averages between about 40 inches and 100 inches (101.6 cm and 254 cm), corroborating the observations described above. Normal annual precipitation was among the most important variables in the Maxent model, as indicated by jackknifing results and contribution to model gain, and likely strongly contributed to the eastern portions of Washington and Oregon having low probabilities of being suitable habitat.

In order to classify model results into suitable and unsuitable Puget Oregonian habitat, we used the threshold which maximizes the sum of model sensitivity and specificity. The suitable habitat identified by the model is located west of the Cascade Mountain Range, throughout the Puget Trough and Willamette Valley. The spatial extent of the model only covered the states of Washington and Oregon, but the modeled swath of suitable habitat likely extends into British Columbia and northern California, coinciding closely with the range of the bigleaf maple. The most notable deviations of the habitat model from the range of the bigleaf maple are the western side of the Olympic Peninsula and the Oregon Coast Range, which this model identified as unsuitable habitat despite being part of the bigleaf maple range. The exclusion of these areas is most likely due to the amounts of normal annual precipitation in these areas, in excess 100 inches (254 cm) per year, which are higher than those observed at existing species records. However, while the species are absent in these areas due to the geologic history of the region, as described in the Species Information section, they likely provide suitable habitat conditions.

Current Range of the Puget Oregonian

12 For the purpose of determining the current range of the snail, we have limited the extent of suitable habitat both in the north and the south. At the northern end, the extent of the range will be cut off at the Canadian border. There have been three records of Puget Oregonian presence between 1850 and 1905, including one near the southernmost point of Vancouver Island; however, the species has not been observed in Canada since then, and has been considered extirpated from the country since 2002 (Ovaska and Forsyth 2002, p. iii; BC Invertebrates Recovery Team 2008, p. 1,). At the southern end, we limited the range to the southern boundary of the Northern Oregon Coastal (HUC 171002) and Willamette (HUC 170900) basins. Due to the snail’s status as a “Survey and Manage” species under the NWFP, the USFS, and the BLM survey or have surveyed for the species before projects are implemented. Despite extensive surveys, the snail has never been found south of

Miles-===--==---- the Northern Oregon Coastal and Willamette watersheds, Species Range 0 40 80 160 Habitat Suitability so the species is unlikely to be found there, probably due Unsuitable to historic geologic events combined with the snail’s • suitable limited dispersal abilities (Figure 6) (Frest and Johannes □ State Boundaries 2000 p. 186; Frest and Johannes 2005, pp. 49- 58). • Species Occurrence Records The range map we developed for the Puget Oregonian Figure 6: Current range of the Puget Oregonian in Washington and Oregon. The Species Occurrence Records using a Maxent model and some historic information shown include both current and historic species records. shares many similarities to existing range maps, but has key differences from some of them. The range description inclu ded in the ISSSSP report (Foltz Jordan and Hoffman Black 2015, 31) identifies a very similar range to our range maps herein, as many data records used to describe the range were also used to develop the Maxent model. The range maps by other sources (Vagvolgyi 1968, p. 225; BC Invertebrates Recovery Team 2008, p. 3; Ovaska and Forsyth 2002, p. 6) encompass a narrower swath of land, as measured from east to west, and do not extend as far south into the Willamette Valley, but include the southernmost portion of Vancouver Island. Burke (2013, p. 158) identifies a similar range for the species as these sources, but does not include any Canadian locations, similar to our model and in the status from British Columbia (BC Invertebrates Recovery Team 2008, p. 1; Ovaska and Forsyth 2002, p. iii).

Throughout this SSA it is important to distinguish what we understand about the species range and distribution from what we understand about this species abundance. Despite the relatively wide distribution of this species west of the Cascade Mountain Crest in the Pacific Northwest its abundance appears to be very low outside of the Cispus Watershed in Lewis and Skamania Counties, Washington, where roughly 90 percent of all recent observations have occurred since the species was first described (Jordan Foltz and Hoffman Black 2015, pp. 10- 11; USDA 2011, p. 20).

HABITAT

General Habitat

Several aspects of Puget Oregonian habitat are discussed in the range and distribution section above. These include elevation maximums, precipitation (moisture) levels, density of vegetation, and presence of bigleaf maple. Those variables are important to understanding the broad ecological setting in which Puget Oregonian are found. Overall, terrestrial gastropods seem to be more common or even dependent on deciduous broadleaf 13 forests, or forests with a mixed conifer – broadleaf composition and in the PNW require relatively undisturbed forest floor or understory cover (Abele 2010, pp. 36-39; Frest and Johannes 1993, p. 3; Foltz Jordan and Black 2012, p. 5). At the landscape level, the snail occurs in what remains of Oregon and Washington’s mature to late successional forests from the Cascade Crest west at low to mid elevations on low to shallow gradients (Burke 1999 p. 5; Foltz Jordan and Black 2015, pp. 11-12). Generally the Puget Oregonian is described as a mature or old growth forest associated species that often, but not exclusively, occurs adjacent to riparian areas above flood-prone floodplains (Foltz Jordan and Black 2015, p. 11; Burke 1999, p. 5). This description seems to describe benched river terraces up to and including the lower valley slopes where bigleaf maple often occurs (Sudsworth 1967, p. 389). The Oregon Forestsnail (Allogona townsendiana), a PNW native snail in the same family, is reported to be generally common in moist coastal forests and more restricted to riparian habitat further inland (Kozloff 1976, p. 81). In general, terrestrial mollusk diversity and abundance tend to be positively correlated with the percentage of hardwood cover, ferns, woody shrubs, seeps, and wetlands (Foster and Ziegltrum p.254; Frest and Johannes 1993, p. 3; Foltz Jordan and Black 2012, p. 5). The Puget Oregonian may follow a similar distribution pattern, but that pattern is unproven because of limited demographic and distribution information and the loss of potential historical habitat in the central Puget Sound area to development.

At a finer scale the Puget Oregonian is almost always associated with mature forest with high levels of canopy cover (typically 70 percent or greater) composed primarily or partially of bigleaf maple often mixed with other hardwoods (red alder (Alnus rubra) or black cottonwood (Poplus balsamifera) and conifers (Burke 1999 p. 5; Foltz Jordan and Hoffman Black 2015, pp. 11-12) (Figure 7). Younger, smaller diameter bigleaf maple stands, or conifer stands with a high numbers of smaller individual bigleaf maple also can also support the species (Burke 1999, p. 6).

Figure 7: Stand level habitat of the Puget Oregonian includes overstory of mature bigleaf maple and understory with pieces of decaying large woody debris, swordfern skirts, and deep leaf and forest litter maintain a moist and cool microhabitat required for snails to be in an active state. Under the dappled canopy of the maple or mixed maple canopy the snail can be found under logs and other hardwood debris, swordfern skirts, leaf litter and/or talus, and around seeps and springs (Burke 1999, p. 5; Foltz

14 Jordan and Hoffman Black 2015, pp. 11-12) (Figures 7 and 8). Young individuals have been reported under mosses growing on the trunks of bigleaf maple as well (Foltz Jordan and Hoffman Black 2015, p. 12). The specific environmental conditions that support the Puget Oregonian in its habitat are unknown. Appropriate temperature, moisture levels, and perhaps soil composition and chemistry are important for the snail. Broadly, we understand that high overstory cover (≥ 70 percent) shades the forest floor and reduces evaporation rates where the snails live (Burke 1999, p. 5; Foltz Jordan and Hoffman Black 2015, p. 12). In areas that are naturally wetter, lower canopy cover has been reported and it is thought increased moisture offsets the reduced level of shading (Foltz Jordan and Hoffman Black 2015, p. 12). Forest litter, large woody debris, swordfern skirts, and other elements create microclimates with cooler, moister conditions that promote snail activity (Figures 4a - c). Bigleaf maples produce one of the highest litter fall by weight in the western states (Peterson 1999, p. 23). Additionally, bigleaf maple litter contains high levels of potassium and calcium, and other macro- and micro nutrients (Peterson 1999, p. 23). This combination of high litter fall creating suitable forest floor microhabitat D conditions and higher levels of available calcium, which is needed to Figure 8: A sampling of habitat where Puget Oregonian were located in produce a snail’s shell, may be a part of May 2019. A. A moss covered decaying maple log under a canopy of bigleaf maple. B. matrix of leaf litter, sword fern, and branches. C. An the reason the Puget Oregonian is undecayed tree stem with good ground contact. D. A nearly homogenous strongly associated with the bigleaf stand of bigleaf maple growing in relatively gentle terrain. maple. If that is true, it is curious why other large snails such and Monadenia, Allogona, Ancotrema species, while found in association with Puget Oregonian and maple, are not as uniquely tied to the bigleaf maple. While we do not definitively know the reason for the association between the bigleaf maple and the Puget Oregonian there is a list of compelling structural, physical, chemical, and ecological reasons for the association. Habitat Components Bigleaf maple trees, as noted above, are highly associated with occurrences of the Puget Oregonian. Bigleaf maple can live upwards of 250 years, grow to 115 ft (35 meters)tall at maturity, and the stem density of the immediate stand can influence the form of the bigleaf maple crown such that the tree has increasingly broad canopy crown (decurrent crown form) as stand density decreases (Pojar and Mackinnon 1994, p. 45; Peterson 15 1999, p. 13). The bigleaf maple is the largest and longest living of the three maple species found in the range of the Puget Oregonian (Sudsworth 1967, pp. 386-387). The other two species being vine maple (A. circinatum) and Douglas maple (A. glabrum) are much smaller, comparatively produce smaller diameter and volume of large woody debris, and lesser volume of leaf litter annually. Bigleaf maple is found from sea level to about 3,000 ft (900 m) in Washington and Oregon. The tree grows on a relatively wide range of soils but best growth occurs on fluvial sites and at the base of colluvial slopes. Growth is best where soils are moist, either from seepage or on fluvial sites along streambanks, but it is not as tolerant of flooding as most other native forest trees (Peterson 1999, p. 9, 16-17). During growth, bigleaf maple absorbs large quantities of nutrients, much of which are returned to the forest floor in its litter. Weight of litter fall and litter nutrient content (for every macronutrient and most micronutrients) were significantly greater under maple than under Douglas-fir (Pseudotsuga menziesii) on sites studied in western Oregon (Fried et al. 1990, p. 259). Litter fall weights are greater under maple than under Douglas-fir, and maple leaves and litter contain higher concentrations of potassium, calcium, and other macro- and micro-nutrients (Peterson 1999, p. 23 referencing Tarrant et al. 1951).).

The bigleaf maple has several features indicating it may be an autogenic ecosystem engineer (organism that creates habitat for species through its structure, growth, and death) for Puget Oregonian and other forest species (Jones et al. 1994 pp. 374 -376). Several components of the bigleaf maple are thought to be especially important for supporting resource and life history needs of the Puget Oregonian as discussed below. Canopy: As stated above, maple canopies can be variable in dimension depending on stand density. Regardless of size, the canopy has a shading, cooling effect on the forest floor where the Puget Oregonian resides, reducing evaporation from heat and wind, moderating moisture loss, and supporting an epiphytic community of lichen, plants, bryophytes, and fungi in its branches. The canopy also is the source of leaf litter and many of the large branches which contribute to the large woody debris on the forest floor. Leaf litter: Big leaf maples produce a high volume of leaf and seed/samara (the winged, “helicopter part”) litter. This leaf and seed litter absorbs and retains moisture, acts as a bases for fungal decomposers (thought to be a food source), insulates from moisture loss and temperature extremes reducing the severity and intensity of freezing and desiccation, and provides refugia from avian and terrestrial predators. Large woody debris: Terrestrial gastropods rely on woody debris for food, shelter, and as a site for breeding, and fungal food base (Harmon et al 1986. p. 235, Frest and Johannes 1993, p. 3). Large woody debris moderates temperature levels and retains moisture creating daily and season microhabitat refuges for terrestrial gastropods (Harmon et al 1986. p. 235, Frest and Johannes 1993, p. 3). Descriptions of the Puget Oregonian’s habitat and ecology are tied to large woody debris contributed by bigleaf maples and other native tree species (Frest and Johannes 1993, p. 30; Frest and Johannes 1995, p. 228; Duncan et al, 2003, p. 41; Forsyth 2004, p.153). After the removal of large woody debris from an occupied site the species became very difficult to find (Foltz Jordan and Hoffman Black 2015, p. 21) likely indicating a decrease in species abundance.

Swordfern skirts: There is little primary literature that discusses the ecological role and microhabitat conditions provided by the “skirts.” Skirts are defined by Marco (2011, p. 1) as the dead vegetative matter, such as the lower leaves or fronds, which instead of falling off remain and form "skirts" of dead vegetative cover. Haggard (2000, p. 60) also describes this microhabitat as fern “enclosures” in work on red-legged frogs (Rana aurora). The Puget Oregonian is well documented in its use of swordfern skirts during the active season for this species (Burke 1999, p. 5; Foltz Jordan and Hoffman Black 2015, pp. 12-13). These skirts can provide cover from predators, thermal refuge during summer and winter seasons, and a moist microsite that persists later in the year than other areas on the forest floor (Haggard 2000, p. 60; Schuett-Hames 2004, pp. 38, 63, 73). Swordfern skirts, much like large woody debris and leaf litter, can provide important microhabitat that provides

16 moisture, temperature moderation, and refugia from predators as the ambient environmental conditions diminish for snail activity. Talus: Talus habitat results from the gradual accumulation of weathered rock fragments at the base of cliffs or other steep slopes (Maser et al., in Thomas 1979, p. 99). Large, deep, and older talus accumulations are described as being most important to wildlife, likely because of the stability of the talus field, establishment of a vegetative community, and size of habitat it provides (Maser et al., in Thomas 1979, p. 99). Individual talus slopes can be variable in rock size, aspect, and in the amount and type of vegetation present creating a broad range of thermal and moisture regimes for wildlife (Herrington in Szaro et al., 1998 p. 216). Herrington (in Szaro et al., 1998, pp. 219-220) identified three categories of use of talus slopes for reptiles and amphibians: 1) talus restricted species, 2) species using talus as refuge from environmental conditions, and 3) reproductive site. The Puget Oregonian has been noted to co-occur with plethodontid salamanders in talus habitat (Crisafulli et al. 2008, p. 13). The most analogous talus use for terrestrial gastropods is likely number two above where the species uses larger, older, moist, mixed, bigleaf maple-forested talus areas, as a refuge from environmental extremes. Talus accumulations represent and occasional habitat type for the Puget Oregonian (Foltz Jordan and Hoffman Black 2015, p. 13).

Spring-seeps adjacent habitat: Appropriate moisture levels are important for terrestrial gastropods to maintain (Hyman 1967, vol 6, pp. 626-627). Many PNW terrestrial gastropods thrive in lowland to middle elevation moist (often riparian) forests often in areas around perennial moist areas such as springs, bogs, or marshes (Frest and Johannes 1993, p. 3). Puget Oregonians are among those often found in association with spring and seeps (Frest and Johannes 1995, p. 229; Burke 1999, p. 6; Foltz Jordan and Hoffman Black 2015, p. 11). Further evidence that moisture is critical to this species is highlighted by the Puget Oregonian inhabiting stands with lower level of canopy cover if the forest floor is wetter (Burke 1999, p.6). Forested springs and seeps represent an important habitat component for the Puget Oregonian.

LIFE HISTORY OF THE PUGET OREGONIAN

There is little specific information about the biology of the Puget Oregonian or any member of the Cryptomastix genus. However there are a series of investigations into the ecology and biology of another PNW Polygyrid snail Allogona townsendiana, The Oregon forestsnail (Steensma et al. 2009, entire; Edworthy et al. 2012, entire). Information about the Oregon forestsnail is used sparingly as a surrogate for the Puget Oregonian in the SSA. However caution and care should be used when using surrogates, especially when comparing little studied species, as each species has its own unique characteristics and adaptations. The life cycle of any animal is the period involving the succession of one generation to the next through reproduction. The life cycle, a terrestrial gastropod can be broken into two states – the active or moving state and the dormant or roosting state (Barker 2001, p. 447). The active state is when most of the gastropods environmental needs are met for feeding, growing, movement/dispersal, mating, and egg deposition. The dormant or roosting state takes place during periods of time when environmental conditions are not conductive to carrying out life history activities and the gastropod sequesters itself in a benign refugia or environment; in snails this is often a mucus plugged aperture, where the snail stays sequestered until conditions improve (Hyman 1967, vol 6, p. 627; Barker 2001, p. 74; Burke, 2013, p. 13). This period can be over the course of a day until the dew level and air temperature drop or over the course of months or perhaps even years in some species (Baker 1958, p. 141). The Puget Oregonian is thought to be most active in spring and fall when cooler temperature and moisture promote snail activity. This species, as is true for most terrestrial gastropods, is most active during crepuscular and nocturnal hours and perhaps diurnal during periods of cool temperature and rain (Cook in Barker 2001, pp. 461-464; Tom Kogut 2019, pers. comm.) Longevity of the Puget Oregonian Life spans of terrestrial gastropods range from several months to 19 years (Heller in Barker 2001, p. 436). Species of the Polygridae family have been found to live for 8 or more years in captivity after reaching maturity 17 (Foltz Jordan and Hoffman Black 2015, p. 8). The lifespan of the Puget Oregonian is unknown and to the best of our knowledge no individuals have been tracked across years. Generally Puget Oregonians are assumed to live for several years (Foltz Jordan and Hoffman Black 2015, p. 8). Heller (1990, p. 264) broadly correlated shell characteristics in terrestrial gastropods with longevity and described two general categories: short-lived and long-lived. Short-lived species are those living up to 2 years or those living longer but reproducing during only one season, and typically having semitransparent shells or lacking shells. Long-lived species are those species that live for more than two years and breed over at least two seasons, having opaque shells (Heller 1990, pp. 264-266). As the Puget Oregonian has a juvenile form that likely lasts at least a year, is opaquely shelled, and lives in a cryptic habitat, we can assume its life span exceeds two years.

In its most basic components, the snail life cycle proceeds through the following stages (Figure 9): 1. Eggs hatch at deposition site and disperse. 2. Juveniles forage, grow, develop, disperse and possibly establish a “home range”. 3. Individuals reach sexual maturity. As this species is likely hermaphroditic we speculatively assume both sets of sexual organs and gametes mature at the same time, but one may mature before the other (Tomiyama, 1993 referenced in Heller in Barker 2001, p. 416). It is assumed that external maturity can be evaluated by development of the reflected apertural lip of the shell and parietal tooth (Fig 2e). 4. At maturity individuals locate a partner and reproduce (or potentially self-fertilize). 5. After mating, sperm may be stored for some time before ova are fertilized. Afterwards the individual gestates an unknown number of eggs for an unknown period of time. 6. Individual deposits eggs in a suitable location, likely under large woody debris, under swordfern, or soil under or near bigleaf maples. After hatching, juveniles disperse through suitable habitat. Although the life history diagram is presented in a cyclical fashion there is little information to understand the duration or seasonal periods in which each life history event happens. Additional information about each life history stage will be discussed below.

18 2 ~ ,,,,,. I 4 ~ I («;".) ?I Poss;bility of self-fertiU,aUoo 1 ~~\i:

-~ \II~ W11/,. ' -:.-.\\\\) ~,.,...,..,~~~ - 6

Figure 9: Basic life cycle of the Puget Oregonian. A list correlating the lifecycle to each stage can be found above. Briefly: (1) eggs hatch, followed by (2) juvenile development into (3) a sexually mature individual. Reproduction is achieved through the (4) mating of 2 or more individuals (although self-fertilization is possible), (5) gestation and (6) deposition of eggs

Reproductive and Developmental Life History

1. Eggs of the Puget Oregonian from Deposition to Hatching:

There are no known reports about the number, size, or duration of development of Cryptomastix species, but we do assume that they lay eggs (Frest and Johannes 1993, p.6; Burke et al 1999, p. 5; Burke 1999, p. 4; Foltz Jordan and Hoffman Black 2015, p. 8). General information about projected egg size, clutch/cluster number, and duration of development for this species is discussed below.

Egg size has been found to be correlated to the size of the snail (Heller in Barker 2001, p. 425). For a mature Puget Oregonian, reported to have a 20-24 mm wide shell, eggs are anticipated to be approximately 2.4 mm in diameter based on reported sizes from other species (Heller in Barker 2001, pp. 425-428). The size of an egg is important to understand because of the relationship between increasing hatchling size and hatchling survival (Heller in Barker 2001, p. 425). Smaller eggs are at a higher risk of failure from environmental stressors (heat, desiccation) than those from larger native and non-native gastropods such as Arion rufus. The number of eggs in a clutch or cluster of eggs can vary individually based on the size and age of the individual. In general, within a species the larger and older individuals have more eggs.

19 However they are deposited terrestrial gastropod eggs are most commonly deposited in moist soil; drought or dry soil is the most common mortality factor for their eggs (Heller in Barker 2001, p. 420).

Other PNW Pulmonate snails are known to deposit eggs in spring and hatch 8 to 9 weeks after deposition (Frest and Johannes 1993, p.6; Steensma et al. 2009, p. 237); whether this incubation period is similar for the Puget Oregonian is unknown. Parental care of eggs or of young hatchlings is not known to occur among terrestrial gastropods (Foltz Jordan and Hoffman Black 2015, p. 8; Baur 1994, p 10)

2. The Juvenile and Mature Puget Oregonian After hatching, and outside of reproductive efforts, most of the Puget Oregonian’s effort is spent acquiring food and water resources for growth and subsistence, locating roosting, aestivation, or brumation sites (shelter), and perhaps dispersal. These activates are further described in the discussion of the active and dormant life history periods below. As an individual acquires resources it grows larger, increasing shell dimensions. At maturity after a period of one or more years, it gains a recurved apertural lip and parietal tooth (Figure 2b). In Allogona townsendiana sexual maturity was found to occur at 2 to 3 years and around 20 to 23 mm in shell diameter (Steensma et al. 2009, p. 239). As in Allogona townsendiana, development of secondary sexual characteristics (i.e., the recurved lip and tooth) in combination with larger shell size are considered by species experts to indicate mature individuals capable of reproductive behavior and producing viable offspring (Burke 1999 p. 3; Foltz Jordan and Hoffman Black 2015, p. 8). 3. Mating, Sperm Storage, Fertilization, and Gestation

As noted above, this species is hermaphroditic as are most snails. Therefore, when snails copulate the opportunity exists for each individual’s eggs to become fertilized. In some Pulmonates hormones released through the insertion of a “love dart” prior to copulation stimulate the receiving individual’s hormones to function as the female or male sex. A defining characteristic of the Polygyrid is the lack of the “love dart.” We do not have information to suggest that either individual takes on a dominant sex role during copulation (Pilsbry 1940, p. 576; Davison and Mordant, 2007, p175). It is unclear at this time if copulation always, sometimes, or never results in fertile zygotes in each individual. Another reproductive option for hermaphrodites is self- fertilization, which of course does not require a partner. Self-fertilization is a common but not a universal phenomenon in Stylommatophora, with the frequency of self-fertilization varying greatly among species (Barker 2001, p. 98).

Frest and Johannes (1993, p. 6) discuss breeding for this genus of snail in April – June. Tom Kogut (2019, pers. comm.) stated the only observation of breeding he had encountered in this species was in October. While both spring and fall in the PNW are similar in temperature and moisture availability, and likely the most active period for the snail, this temporal dichotomy in identification of breeding periods lends itself to question how long the species produces eggs, when are eggs laid, and if internal fertilization of eggs can be delayed.

Steensma and others’ (2009, p. 235-236) work on Allogona townsendiana reports breeding most often takes aggregations of eight to fourteen snails, but was observed once as a pair, in the Fraser Valley of British Columbia. These events took place between February and June with soil temperature from 9.9° to 13° C, soil moisture between 30 to 37 percent, with neutral pH (Steensma et al. 2009, p. 236). These observations of a species of the same family breeding in the spring, support Frest and Johannes (1995, p. 5) statement that spring is the typical mating season for Polygyrid snails in the PNW. Although the only known observation of Puget Oregonian mating is from the fall, it seems reasonable, at this time, to consider that anytime the Puget Oregonian is active it also has the opportunity to mate.

4 Egg Deposition. At this time, there is no empirical information about where or when eggs are deposited by this species. However, there are several possibilities. One is that Puget Oregonian eggs are laid in a clusters in a shallow

20 excavated nest, as observed in other PNW Polygyrid snails (Steensma et al. 2009, p. 236). Other possibilities are that they are laid in or under large woody debris, or that they are laid within interstitial spaces in the substrate where the snails shelter. To date, no observations of Puget Oregonian nesting or nests have been documented. Due to similar oviposition, it seems reasonable to conclude that their eggs are deposited under a cover object or in the upper areas of the soil where conditions are favorable.

The Active State of the Puget Oregonian

Most land snails are only active in spring, late summer, and fall when temperature is cool but not too cold and moisture levels are wet. The available moisture levels at other times of years are most suitable after rain and at night, while during dry hot summer periods snails become inactive (Hyman 1967, Vol 6, pp. 626-628; Cook in Barker 2001, pp. 456-457; Forsyth 2004, pp. 6-7). The active season of the Puget Oregonian snail is thought to be primarily in the spring and fall when moist cool conditions dominate the lush forestlands of the PNW (Forsyth 2004, p. 7). This is time of year in the PNW when many moisture dependent terrestrial organisms are most prominent, including gastropods, mushrooms, moss, lichens, and amphibians (Frest and Johannes 1993, p. 33). However, some PNW terrestrial mollusk activity has been observed under cover during the winter, even under a layer of snow (pers. comm. Foster 2019). Snails are active in a temperature range of approximately 1 to 19°C with appropriate levels of moisture available (Cook in Barker 2001, p. 459). Terrestrial mollusks in conifer-forests of western Washington were found to be most active between 4° and 7°C (39 to 45°F) (Foster and Ziegltrum 2017, p. 250). We assume that terrestrial gastropods in the temperate forests of the PNW are likely intolerant of freezing temperatures and individuals seek refugia from it (Frest and Johannes 1993, p. 33; Ansart et al. 2001, p. 186). These refugia are likely under logs, swordferns, and in thick litter, among moist talus field or in the soil. These provide short or long-term refuge from unsuitable environmental conditions and are likely used to extend the active period for snails on a daily and seasonal scale. The daily components of gastropod activity are broken into three periods (Cook in Barker 2001, p. 456):

1. Emergence: most often thought to occur at crepuscular hours, when animals become active and leave their roosts; 2. Excursion: animals make an excursion during which feeding, drinking and courtship are interspersed with periods of rest; 3. And roosting: animals locate a suitable resting site, often by homing to previously occupied sites, and come to rest.

The Roosting and Dormant State of the Puget Oregonian.

Roosting is the daily period when a snail is not active and stays within its shell. This may occur when daily ambient conditions are undesirable such as being too hot, cold, or dry. Some snails have consistent roosting areas that they return to on a regular basis and journey to and from frequently. These regular roosting areas likely provide refuge from predators and environmental conditions. Dominancy or Aestivation During summer months when it is dry and warm, Puget Oregonian are not found on the soil surface, or within the organic soil layers, including under large woody debris and or swordfern skirts. Very little is known where Puget Oregonian goes to survive the summer heat and drought. Where they estivate in summer is unknown, but considering the “large” size of their shell it is unlikely they can move through the soil directly. It is likely Puget Oregonians rely on refuge habitat in downed, decayed LWD or small fissures and openings in wood or talus. When the snail is inactive for a period of time it creates epiphragm, or mucus plug in the aperture, to seal the shell and reduce rate of water loss during dormancy (Cook in Barker 2001, p. 455; Forsyth 2004, p. 180).

21 LIFE ON THE FOREST FLOOR:

Movement and dispersal: “A bag of cold water that cannot even move unless it leaks should not be able to survive outside a bog” – Rollo and Wellington 1977. (And yet move they do!) Movement, Riding a Wave of Mucus:

Terrestrial Gastropods “swim” on a thin layer of self-produced mucus secreted through a suprapedal gland, or pore, near the anterior end of the foot (Baker 1958, p. 142; Hyman 1967, vol. 6 pp. 561-562). The mucus or “slime” released from the pore makes the track or trail upon which the snail moves via the beating of microscopic cilia, on the sole of the foot. The beating of the cilia momentarily liquefies a portion of the mucus allowing the gastropod to slide forward (Hyman 1967, vol. 6 pp. 628-629; Ruppert et al. 2004, pp. 313-314). “Thus a land snail really swims over the land, along a little river of slime” (Baker 1958 p.142). This mode of locomotion is both slow, measured in cm/minute and requires a high rate of water loss due to constant creation of mucus. Thus the individual snail may choose estivation over looking for mates, foraging, or dispersal during periods of low moisture or high temperature (Hyman 1967, vol 6. p. 644 -645; Cook in Barker 2001, pp. 447, 457, 467-468).

Movements of snails from roosting or home sites are attributed to seeking food, mates, water, winter and summer refuge areas, or excursions to investigate new resources of the above (Cook in Barker 2001, p 466). The longest movement recorded over 3 years for another similar-sized PNW land snail, the Oregon forestsnail, is 32.2 m (Edworthy et al. 2012, p.878). We have noted at least one instance where we believe the Puget Oregonian made a movement of at least 2 m in a year’s time (Le and Waterstrat 2019, pers. obs., p. 2).

Dispersal

Dispersal ability should not be confused with the snail’s range size. Range size is an outcome of dispersal and persistence of the species and is not only determined by the dispersal ability of a species, but also by its ecological tolerance (Hausdorf and Hennig 2003, p. 106). Macroscopic terrestrial snails are relatively poor at dispersal, and generally the larger the snail the poorer it is at dispersing (Hausdorf and Hennig 2003, p. 106; Nekola 2014, p. 233). Below we briefly discuss the two most common modes of terrestrial gastropod dispersal: active and passive dispersal.

Active Dispersal Active dispersal abilities (i.e., crawling/gliding) in medium and large snails, such as the Puget Oregonian, is limited and likely does not greatly exceed 33 ft (10 m) (Edworthy et al. 2012, p. 878-879; Hausdorf and Hennig 2003, p. 106). Many snails are also unable to actively move cross physical or habitat barriers of 328 to 3280 ft (100 to 1,000 m) even over generational times, and narrow barriers, including roads can act as barriers to individual active dispersal (Bauer 1988 p. 251; Schilthuizen and Lombaerts 1994, p. 582; Jordan Foltz and Hoffman Black 2012, pp. 21-22). As noted above, we believe we have evidence of Puget Oregonian active dispersal or movement of a minimum of 2 m in a years’ time (Le and Waterstrat 2019, pers. obs., p. 2).

Passive Dispersal Successful movement via passive means, such as moving on another larger organism like a bird or mammal or being blown by wind is inversely correlated with land snail body mass for two reasons. First, smaller individuals are less likely to become detached from the carrier and second the ability to found a new population via movement of only a single individual appears to be more common in tiny snails such as Vertiginids. To establish a new colony, larger snails, which are more likely to be dependent on sexual reproduction, require contemporaneous movement of at least two individuals or single adult storing sperm (Nekola 2014, p. 233). Movement via water is possible, but there is little documentation about this method of dispersal. In contemporary times passive dispersal of large snails may be more common due to anthropogenic means such as

22 hitching a ride on birds, vehicles, or by individual collection and release (Rees 1965, p 269; Burke 2013, p. 26; Waterstrat 2019, pp. 5-6).

In summary, larger snails are much more likely to experience isolation, restricted gene flow, and allopatric speciation as compared to small snails because they are less likely to disperse passively throughout the range of physical and environmental conditions to which they are adapted. However recent human assisted movements may allow for greater contemporary passive dispersal.

Diet and foraging: Numerous food items have been recorded as part of a terrestrial gastropod’s diet: plants of all developmental stages from seedlings to senescing plants, leaf litter, wood, microbial, fungal biomass, and dead animals (Speiser in Barker 2001, p. 262). What little we know about the diet of Polygridae snails in the PNW, including the Puget Oregonian, indicate that they forage in and on the litter of the forest floor as primary consumers foraging at the scale of the forest ecosystem and primary or secondary consumers at the scale of the soil ecosystem (Brady and Weil 2002, p. 317; Marcot and Vander Heyden 2001, p. 171).

Pilsbry (1940, p. 575) states of the Family Polygridae, "Their food is chiefly the mycelia of fungi." Other species experts indicate their professional judgement is that their food is primarily fungivore or other microorganisms associated with decaying leaf litter such as molds, yeasts, and bacteria (Burke 1999, p. 4; Foltz Jordan and Hoffman Black 2015, p. 9). Juvenile Puget Oregonians will consume lettuce in captivity providing insights that they can consume leafy vegetative material (Burke 1999, p. 4). It has been speculated that one snail species could eat hundreds of species of fungi that are unique to that particular forest (Rundell in Platt 2016). Terrestrial gastropods have been noted as “prominent damaging agents” of bigleaf maple seedlings, but there is no information that Puget Oregonian itself eats their seedlings (Peterson 1999, p. 38). Soil itself and the humic acids and micronutrients that it contains can be important for the growth and development of the species (Speiser in Barker 2001, p. 263).

Although most contemporary literature places terrestrial gastropods as primary consumers, or perhaps secondary consumers if consuming decomposers, it is important to recall the bias of the classic wildlife biologist or ecological researcher’s perceptions of scale when making these generalizations. From the perspective of life on and in the forest floor, that of ganglia of the soil organisms, and in the mind of the soil ecologist, a mere handful of soil may contain billions of individual lifeforms filling countless niches from producer to apex predator (Brady and Weil 2002, p. 6).

Foraging is the manner in which an individual seeks out and allocates time to obtaining food resources. For this species foraging takes place during the active period of its life history primarily thought to occur in spring and fall. This species is active during nocturnal and crepuscular hours and perhaps diurnal during periods of cool temperature and rain making this the time period foraging occurs (Tom Kogut 2019, pers. comm.). It is not completely clear where in its habitat the foraging occurs or if it is entirely opportunistic. It is most likely that the Puget Oregonian uses chemoreceptors to find food and then further investigates the palatability by scrapping up the food item with its radula, bringing it into its mouth and then accepting or rejecting it.

General Foraging Theory states that there is a cost benefit relationship for most organisms when selecting optimal food times and the time required to obtain the food item. Locomotion of terrestrial gastropods is slow and costly because a mucus trail is secreted along the entire path covered by the animal (Denny 1980, entire). This mode of locomotion is likely the reason why gastropods adopted a generalized foraging strategy which does not rely on large movements to locate specific food resources (Speiser in Barker 2001, p. 262). Speiser (in Barker 2001, p. 264) states the availability and accessibility, as well as by the nutritional needs of the gastropod, may be subject to seasonal changes and as a consequence, gastropod diets may vary greatly over the season.

While the majority of this discussion focuses on the snail diet (vegetative and fungal items), it is important to emphasize that obtaining water and staying hydrated, for locomotion, respiration, and physiological processes, is perhaps a greater priority to the individual than obtaining food resources on a day-to-day basis. We anticipate 23 that foraging occurs most often in the active season and is reduced to more limited forays as environmental conditions become hotter and drier or too cold for the physiological limits of the snail.

Respiration: Pulmonate snails are unique in the gastropods because they “breathe air” while most others use gills to extract oxygen from the water. The Pulmonate have transformed their mantle cavity into a pulmonary sac or pallial cavity absent the ctenidium (gill) found in most other gastropod classes (Hyman 1967, vol. 6 pp. 558, 566) (Figure 1). The moist sac or cavity has many capillary vessels where gas exchange takes place (Barker 2001, p. 50). The pulmonary sac has a single exterior opening, or pneumostome, typically found on the dextral (right) side of the snail through which gases are exchanged (Hyman 1967, vol. 6 p. 558; Barker 2001, p. 50) (Figure 1). Having an enclosed cavity with a small opening reduces the water loss from the cavity and is at least partly responsible for the radiation of gastropods species into terrestrial habitats and allows respiration during periods of low environmental humidity (Barker 2001, pp. 50-51).

SUMMARY OF INDIVIDUAL, POPULATION, AND SPECIES NEEDS

In the Sections above we discuss the taxonomy, range, distribution, habitat, and life history of the Puget Oregonian or our assumptions about its life history needs from other related gastropod families. Below we use that information to define resource needs of the species and describe the current condition of the species based on the SSA framework. At a broad level Foltz Jordan and Hoffman Black (2012, p. 8) summarize the resource needs of Pulmonate gastropods by defining them as limited-mobility organisms with highly specific habitat requirements. Gastropod survival, abundance, and spatial distribution in forest systems is influenced by both local conditions (e.g., microclimate, soil chemistry, vegetation, and available food and refugia) and landscape conditions (e.g., forest type, age, and size) (Foltz Jordan and Hoffman Black 2012, p.8). We evaluate the individual, population, and species needs of the Puget Oregonian in terms of the resource needs that are necessary to complete each stage of the life cycle, including eggs, hatchlings, juveniles, and adults. Broadly, all life stages of the Puget Oregonian are dependent upon the suitably moist, thermally appropriate microhabitats, and dormant season refugia with mature forest stands containing at least some bigleaf maple components.

Individual needs: There is little information available about the specific requirements of individual Puget Oregonians. We do understand forest stand-level physical and biological habitat elements and how they support different stages of the snail’s life history (Table 5). Undoubtedly there are many complex subtleties in an individual snail’s needs for egg incubation, nutrition, growth, and reproduction that we do not understand or address in this document. Despite the advances in our understanding of the species distribution since implementation of the NWFP survey and manage program (Burke 1999, 7; Johannes 2012, p. 9), information about the species’ natural history remains sparse.

24 Table 5: Individual Resource Needs of the Puget Oregonian. Active stage (A), Dormant stage (D). Eggs simply develop without additional needs and will be categorized as D Resources Eggs Juvenile Adult Big leaf Maple/mixed maple forest D A, D A, D Moisture (precipitation >35 in/year - unknown A A springs) temperature 0 - 19° C unknown A A Forest floor habitat components (LWD/Leaf D A, D A, D Litter/Swordfern, talus) Appropriate Pedosphere (primarily Soil 'O' & 'A' horizons) conditions: chemistry/type, D D D interstitial space moisture Food Resources N/A A A

Population needs

We have little information to evaluate what a population of Puget Oregonian is or needs. In the 1990’s Puget Oregonian sites were documented primary in the eastern central Puget Sound and considered small with no examples of large numbers have been reported (Frest and Johannes 1995, p 229). Since that time most survey efforts have been concentrated on federal lands and the current center of the range and greatest abundance is the Cispus River Watershed (Jordan Foltz and Hoffman Black 2015, p. 9).

There are very useful studies investigating populations of Oregon forestsnails, a species in the same family of Polygyrid snails, which could be used as a surrogate for information about Puget Oregonians. However, it is the authors’ opinion that there is a risk in using surrogates for this species as there are identified differences between Puget Oregonians and other PNW snails in that are not fully understood (Burke 1999. p. 7). The Oregon forestsnails studied had a population size of 7 to 47 individuals as defined in 24 square meter study plots and had a mean snail density of 1 snail per square meter (Steensma et al. 2009, p. 335). In 2019, Service and USFS staff found two Puget s under swordfern skirts in a 2 meter squared area in 5 minutes and 4 individuals in a 10 meter by 1 meter transect survey (Le and Waterstrat 2019, pers. obs.); though there was no information indicating these detections were “populations,” the snails’ close proximity to each other makes it likely that they have the opportunity to interact sexually.

Most observations, outside of the Cispus Watershed, are of single or small number of individuals with few exceptions. In 2019, Service staff, USFS staff, and Tom Kogut (USFS retired) revisited a prior survey plot in the Cispus watershed and found both juvenile and mature Puget Oregonians 19 years after they were first identified during a Survey and Manage inventory. This area was subject to intense and severe fire starting in 1902 through 1918 and since that time was replanted and is reaching a mature forest stage (M. Lewis 1939, unpublished report; J. Donahey 2019, pers. comm.). This observation demonstrates Puget Oregonian occupancy and reproduction 80 years after stand replacing wildfires. If we make the assumption that there was no, or very limited, immigration into this survey plot then we can assume there is a persistent, reproductive population within the Cispus Watershed that has survived catastrophic events (wildfire, severe forest practices, road building, etc.).

Other areas where there are observations of Puget Oregonians over several years, but without empirical reproductive evidence, include the 11 acre (4.5 hectare) Crystal Springs Park in Tukwila, Washington (Johannes 2017a, p 1; Johannes 2017b, p 25).).

Given multiple observational records of only a single specimen, perhaps one of the most confounding issues is the lack of information regarding whether this and other Cryptomastix species are hermaphrodites and capable

25 of self-fertilization. However, certainly a single, self-reproducing individual snail is not indicative of a resilient population especially for larger snails (en sensu Nekola 2014, p. 233). Observation of self-fertilization and oviposition of viable eggs and the conditions under which they occur, in a natural setting or laboratory, would be very useful to understanding the ability of this species to recovery from catastrophic or stochastic events.

Despite the number of observations of the species we can state little about the needs of an individual population of Puget Oregonians, beyond the requirement for areas of connected bigleaf maple or mixed bigleaf maple forests with appropriate understory habitat and environmental conditions as discussed in the “Habitat” Section.

Species needs We evaluate the species’ needs in terms of the resources and/or the circumstances that support the redundancy and representation of the species. The viability of the Puget Oregonian is supported by having multiple (Redundancy), self-sustaining (Resiliency) resiliency units distributed throughout the geographical extent of its range (Representation).

FACTORS INFLUENCING THE SPECIES

As stated above there is little specific information about the individual needs a Puget Oregonians or a population of Puget Oregonians beyond the biotic and abiotic habitat conditions, or ecological niche, in which they are found (Burke 1999, pp. 4-5; Foltz Jordan and Hoffman Black 2015, pp. 8-9). We have attempted to synthesize in general terms the known resources that contribute to the resiliency of the species (Figure 10). In addition we describe sources of stressors we understand to the resources on which the resilience and overall viability of the Puget Oregonian depends.

Appropriate soil (pedosphere) Pacific Northwest Maritime Climate, Phys,cal Geology, and contiguous Mature Forest

forested talus fi elds

Figure 10: Known resource needs of Puget Oregonian individuals and populations. Purple boxes indicate regional drivers, orange resources are overarching niche components, yellow microhabitat components, and grey life history needs, all of which contribute to the species viability.

The Puget Oregonian is adapted to the maritime climate of the PNW from the Cascade Crest west to the coastal forests of Washington and Oregon. The seasonal variation in temperature and precipitation is thought to be the

26 primary overall driver of the active and dormant states of the Puget Oregonian. The moderate temperature and moisture of spring awaken the snail from dormancy into its active stage when it can be found under cover on the forest floor. As spring temperature increase and moisture from precipitation and/or snowmelt decreases, the snails become less active on the surface of the soil layers and enter dormancy through the typically dry hot summer. Fall rains cause the snail to emerge from its dormancy for a period before the frosts of fall reduce the periods of activity until the snails return to their refuges for winter hibernation. Mature bigleaf maple and mixed bigleaf maple forest stands are the only places where live Puget Oregonians have be observed and documented. The exact reason for this close association between the Puget Oregonian is unknown, but it is suspected to be related to the large volume of litter fall, fallen branches, and high levels of micronutrients including calcium needed for development of the snail’ shells. Moisture retaining and temperature buffering conditions found in deep leaf litter composting on the forest floor, pieces of LWD, often from the large spreading branches of the maple or other trees, swordfern skirts, and in some cases forested talus slopes and springs provide important microhabitat features for the life history needs of the snail including a rich source of vegetative or fungal food, shelter from predators, and the opportunity to encounter other individuals and copulate. Overall we know little about specific life history needs of the snail and “life in the undergrowth” goes unnoticed and under studied and the opportunities for further research and monitoring are still largely unaddressed (Burke 1999, pp. 4-7, Foltz Jordan and Hoffman Black 2015, p. 5)

Stressors to the species needs:

Forest management

It is important when discussing the distribution and range of the Puget Oregonian to consider the extent of deforestation from early, unregulated logging in the Pacific Northwest. Between 1945 and 1970 approximately 850,000 acres of commercial forest land west of the Cascade Crest was permanently removed with the primary causes being road building and urban and industrial expansion (Bolsinger 1973, p. 4). Overall, early forest management practices and land conversion in the PNW was largely unregulated until the mid-twentieth century when national and state forest practices was codified. In the mid-1990’s prior to the enactment of the NWFP most of the Puget Oregonian’s habitat in the Central Sound area was thought to have been logged and then heavily urbanized (Frest and Johannes 1995. p.229; Mclean and Bolsinger 1997, p. 7). It is generally understood that any modification of habitat that decreases available moisture or increases insolation, like timber harvest, is detrimental to terrestrial gastropods (Frest and Johannes, 1993, pp. 3-4). However, in some instances the impacts of timber harvest may be mitigated by site conditions that retain suitable habitat conditions, such as a north-facing aspect, abundant seeps, and a well-developed and diverse understory (Foster and Ziegltrum 2012, p. 254).

It is possible that significant stands of maple and mix maple forest were removed prior to those regulations. Bigleaf maple is not utilized as a commercial species at a large scale, but may be harvested to a limited extent for the musical instrument and furniture market or illegally felled by poachers. It is also not a desired species, like conifers, for commercial forest owners or state and federal land managers to replant. Thus, it is common for areas containing bigleaf maple as a stand component to have been converted to a more homogenous composition of Douglas fir and Western hemlock or other marketable tree species (Foltz Jordan and Hoffman Black 2015, p.15).

In the late 20th century national and state regulations changed the way timber was harvested with protections for sensitive areas including the designation of riparian corridors, meant to protect water quality and instream habitat conditions, and the retention of some proportion or stem-density of the landscape in uncut timber. 27 While these regulations provided some protections for many habitats and wildlife species there was little emphasis on retaining maple or other hardwoods (Peterson 1999 pp. 3, 55). The exception to this trend is for projects and harvests on federal lands covered by the NWFP subject to Survey and Manage protocols and protections. The largest conservation impact for species associated with old forest was the identification of large areas to be managed as Late-Successional Reserves, where management of these areas was to be for the benefit of these species. In addition, riparian reserves were identified to protect aquatic and old-forest associated species. Including nationally designated Wilderness and other reserve land allocations, over 80% of the Forest Service and BLM land base in the range of the northern spotted owls was classified as “reserve” lands, where management of these lands was to benefit late-successional species, including Puget Oregonians USDI and USDA 1994, p. A-4). Further mitigation to benefit this, and approximately 300 other late- successional associated species, included surveys to be conducted prior to “habitat-disturbing” activities, and any site where these species are detected.

Land Conversion to Agriculture and Development Much of the formerly known range and habitat of the Puget Oregonian has been developed for urbanization, infrastructure, or agriculture by the 1990’s including many of the Central Puget Sound locations (Mclean and Bolsinger 1997, p. 7; Burke 1999, pp. 9, 24). The human population of Washington State alone has increased by an estimated 1 million individuals since 2008 when this species was petitioned (Washington State Dept. of Financial Management 2018, p. 7). While conversion of habitat from mixed conifer broadleaf forests represents a total loss of habitat infrastructure, such as roads, results in habitat fragmentation, populations becoming isolated, a vector for non-native invasive plant and animal establishment, and as a continual source of mortality for snails attempting to disperse across roads (reviewed in Foltz Jordan and Hoffman Black 2012, pp. 20-22). With that population growth comes the need for homes, stores, infrastructure, and resources which have and will continue to alter and eliminate habitat for this species, except where some remnant habitats remain relatively protected from that growth. Despite this development, there are still some protected sites occupied by Puget Oregonian within urban areas such as Crystal Springs Park in Tukwila, Washington, and McAllister Springs, Thurston County, Washington (Johannes 2017, p. 1, BLM 2019). Bigleaf Maple dieback

Since first noted around 2007 (Chadwick et al. 2012, p. 1), maple dieback disease has been implicated in the wide-spread mortality of maple habitat (Ohmdahl et al. 2012, entire). The cause of the disease is unknown and multiple hypothesis have been proposed including positive correlations with higher temperatures, vapor pressure deficits, decreased precipitation, high levels of developed land, low levels of forested or herbaceous land, proximity to paved roads (Betzen 2018, entire), and a pathogen carrier by insect (Donahey 2019, pers. comm.). Regardless of the cause of the die back events, recent surveys have found that it is common and widespread across the range of the Puget Oregonian in Washington and adjoining states (Tyson 2018, p. 29; Betzen 2019, entire). This includes several currently occupied stands in the Upper Cowlitz sub-basin in the heart of the species range (Le and Waterstrat 2019, pers. obs., pp. 1-2). This disease appears to be currently killing large mature maples at both the individual and the stand level. At present, it is uncertain how severely this will impact the current condition of the species, but the significant reduction of bigleaf maples, the foundation of this species’ habitat, seem likely. As this tree is the foundation of Puget Oregonian habitat, it is reasonable to predict that a continued decrease in the abundance of bigleaf maple will negatively affect the snail overtime.

Wildfire

28 Wildfires, of both natural and human origin, have played important roles is shaping the ecosystems in the PNW (Agee 1996, pp. 3, 53). Although part of the ecosystem to which the Puget Oregonian is adapted wildfires are recognized as lethal to snails although the timing, severity, and intensity of an individual fire results in a variety of impacts to individuals and populations (Burke 2013, pp. 18-19; Foltz Jordan and Hoffman Black 2012, pp. 25-28). High-intensity fire poses a threat to the species by removing habitat, directly killing individual snails, and isolating remaining populations (Foltz Jordan and Hoffman Black 2015, p. 17). Typically species that find refuge during the dry periods of the year, such as terrestrial mollusks, when fires are most like to occur are afforded some protection from fire (reviewed in Foltz Jordan and Hoffman Black 2012, pp. 24-25). However terrestrial mollusks are very susceptible to heat and dehydration and have a very limited ability to escape wildfires or disperse into remaining areas of suitable habitat after fires (reviewed in Foltz Jordan and Hoffman Black 2012, pp. 24-25). Puget Oregonians are reported generally lacking from areas where controlled burns were applied after timber harvest (Foltz Jordan and Hoffman Black 2015, p. 17). One accidental fire was reported to result in the mortality of adults in a bigleaf maple patch when burning slash piles (Foltz Jordan and Hoffman Black 2015, p. 17). While we do not know the full impacts of wildfire on individual or populations of Puget Oregonians we do know that wildfires can result in mortality and loss of habitat. Currently wildfires are becoming more common, typically larger in scale, and of higher intensity than the historic fire regime (Dennison et al. 2014, pp. 2930 - 2931).RESILIENCY, REPRESENTATION, AND REDUNDANCY

Analytic Units Resiliency Units As stated earlier, resiliency of a species in a SSA refers to the ability of populations to withstand stochastic events (arising from normal environmental perturbations); we characterize resiliency by using metrics of population health. Resilient populations have sufficient abundance and habitat to withstand stochastic events (wildfire, floods, or drought). The lack of demographic information for Puget Oregonian impeded our ability to formulate demographic metrics for assessing resiliency. Furthermore, we do not have information to determine what a population is beyond a handful of resurvey events indicating that the species still exists in the same survey plot. Inferences about general gastropod fecundity, population areas, and life spans from other studies are summarized in the Species Description discussion above, but without census and demographic information it is difficult to describe a population’s ability to withstand stochastic events. The one exception to this may be in the Upper Cowlitz River Sub-basin of the Cowlitz Valley, which includes the Cispus River Watershed, where the species has been found commonly, but not ubiquitously, in suitable habitat over several decades. Because we lack population and demographic information that would guide us in how to delineate clear populations of Puget Oregonians, we used a geographically and ecologically-based area to define analysis units for this species. We understand that individuals of this species likely have a very limited dispersal range of tens of meters (Backeljau et al. in Barker 2001, pp. 396-397; Edworthy et al. 2012, pp. 878-879; Le and Waterstrat 2019, pers. obs.). Furthermore, there is some evidence that other Pulmonate gastropods distribute themselves along drainages into populations (Arter 1990, p. 997), but we do not know if this holds true for Puget Oregonians as we find them distributed broadly along the Cispus River terrace (Xerces Society 2019). Finally, the accepted potential dispersal for Pulmonate snails is approximately 3281 ft (1 km) (Nature Serve 2019, pp. 5- 6). Therefore, to define resiliency units, we buffered all observations by 1 kilometer, and where and if they overlapped, the resulting areas were merged into a single polygon. Though we use these polygons as Resiliency Units, we recognize that many of the units represent a single observation, and that many populations or occurrences of Puget Oregonians may exist outside the limits of our current knowledge about the species distribution.

29 Using the 1 km distance around known occurrences identified 74 resiliency units (Appendix III). Resiliency units were identified by their representative sub-basin and a numeric code. Twenty-four of the 60 sub-basins within the range of the species have identified resiliency units. There were 17 resiliency units in the Upper Cowlitz sub-basin, 7 in the Middle Columbia and Puget Sound sub-basins, and 6 in the Lewis sub-basin. All other sub-basins had less than 5 resiliency units. The Service did not have sufficient information on our model criteria to analyze current condition status for 80 percent (n=59) of the 74 resiliency units; these units have so little demographic information available and the record of their occurrence is categorized as either “mid- century” or “historical” that they are not considered in our analysis to contribute to the species resiliency. These “unknown” resiliency units will, however, be considered in the overall representation and redundancy of the species viability. The Service had enough information to assess and define a current condition for 15 resiliency units that we assume are currently occupied by Puget Oregonians. These include: Middle Columbia- Hood-4, Middle Columbia-Hood-7, Puget Sound-1, Puget Sound-6, Snoqualmie-1, Snoqualmie-2, Upper Cowlitz-2, Upper Cowlitz-3, Upper Cowlitz-5, Upper Cowlitz-6, Upper Cowlitz-7, Upper Cowlitz-11, Upper Cowlitz-13, Upper Cowlitz-15, and Upper Cowlitz-16. Representative Areas The Service considered the range of the Puget Oregonian to be from the Fraser River Valley lowlands and southern Vancouver Island of British Columbia, Canada, south to the Klamath basin of Southern Oregon - Northern California. As stated earlier in this document, representation, or diversity within the species, is frequently assessed by reviewing the genetic, morphological, behavioral, or environmental diversity found in the species across its range. For this SSA, we did not find, receive, or have knowledge of genetic, morphological, or behavioral differences across the range of the species. Further, the microhabitat conditions, or environmental diversity, appear to be the most important elements of representation in terrestrial gastropods, and these seems to be fairly uniform across the range of the Puget Oregonian. This evaluation left us the option of evaluating representation as an ecological or geographical variable. Below we briefly describe our process for evaluating each resource and our rationale for determining a representative areas for the Puget Oregonian. Focusing on available information we can readily apply across the range of the species, we assessed the usefulness of the following standards as representative areas: United States Geographic Survey’s Watershed Boundary Dataset at different hydrologic levels (USGS 2013, entire), the Environmental Protection Agency’s third and fourth level Ecoregions of the Pacific Northwest (Omernick and Gallant 1986, entire), and Nature Serve’s Element Occurrence data standard (Nature Serve 2002, entire). Using the Watershed Boundary Dataset, the Service evaluated the Cispus River Watershed as a reference representative area because the vast majority (71 percent) of known species occurrences are located in this sub- basin. Starting at the finest resolution of the 12-digit HUC sub-watershed, we used topographic and hydrologic overlays to analyze the distances between occurrences in relation to sub-watershed boundaries. We found that at this scale, the sub-watershed boundaries separated adjacent occurrences while grouping together observations that were separated by larger distances and hydrologic and topographic features that would limit the species’ ability to disperse. The Service repeated this exercise at the watershed (HUC 10) scale, with similar results. At the sub-basin (HUC 8) level, adjacent occurrences were grouped together, making for logical snail representative areas. The Service also coarsely evaluated two other methods of delineating representative areas for this species including evaluation of third and fourth level Pacific Northwest Ecoregional divisions (Omernik and Gallant 1986) and Nature Serve standards for species element occurrences (NatureServe 2002, entire). The Ecoregional divisions, at both the third and fourth level encompassed areas too broad and large for us to reasonably define them as a representative unit for a species with a limited dispersal area. For example the level 4 ecoregions 4a: Western Cascades Lowlands and Valleys and 3d: Valley foothill straddle both sides of the Columbia River.

30 Conversely, we felt Nature Serve’s Element Occurrence standard was too constrained to encompass areas that included areas of suitable habitat conditions, but no records of the species, and may leave out important areas for the species that remain unsurveyed or where the species has not yet been detected. Using the United States Geographic Survey’s Watershed Boundary Dataset HUC 8 (sub-basin) level we divided the Range into 60 sub-basins representative areas. Species records are found within 23 (out of 60) sub-basins in Washington and Oregon (Table 6). Of those 23 sub-basins 19 have records since 1990 and are considered “recent observations” and only 4 contained resiliency units with known condition (these are highlighted in Table 6). Table 6: Sub-basin (HUC 8) representative areas for the Puget Oregonian in Washington and Oregon. Highlighted rows indicate areas that contain resiliency units with known conditions.

Number of Resiliency State Representative Areas (HUC 8) HUC Code Units within

Clackamas 17090011 2

Lower Columbia-Clatskanie 17080003 2 Lower Columbia-Sandy 17080001 1 Lower Willamette 17090012 3 OREGON Tualatin 17090010 1 Yamhill 17090008 4 Deschutes 17110016 1 Duwamish 17110013 2 Hood Canal 17110018 1 Lake Washington 17110012 3 Lewis 17080002 6 Lower Cowlitz 17080005 3 Middle Columbia-Hood 17070105 7 Naches 17030002 1 Nisqually 17110015 4 Puget Sound 17110019 7

WASHINGTON Puyallup 17110014 3 Sauk 17110006 1 Snoqualmie 17110010 2 Stillaguamish 17110008 1 Upper Chehalis 17100103 1 Upper Cowlitz 17080004 17 Upper Yakima I 17030001 I 1

31 Resiliency

Resiliency describes the ability of populations to withstand stochastic events. We can measure resiliency based on metrics of population health and habitat quality and quantity. Highly resilient populations are better able to withstand disturbances such as demographic stochasticity, environmental stochasticity, or the effects of anthropogenic activities. Small populations distributed over a large area are limited in their ability to rebound from stochastic events. To assess resiliency within each unit we developed a series of scoring criteria focused on the limited occurrence information available and our modeled habitat information (Table 7). The first criterion is 1) temporal relevance of occurrence bins on the most recent observation of the species into contemporary or historical categories (defined in Table 7) that reflect the Service’s level of certainty that a population or individual snail may still occur in a resiliency unit. The second criterion focuses on 2) the number of observations of the Puget Oregonian in each resiliency unit. As stated there are no abundance surveys or estimates for this species, only occupancy and observational reports. For the purposes of this assessment, we assume that the more frequently it has been encountered or reported the more likely it is to be more abundant than rare in a resiliency unit. The third and fourth criterion focus on 3) the extent of modeled suitable habitat within and 4) surrounding a resiliency unit. These metrics 3) habitat quality and 4) Potential surrounding habitat connectivity and quality define coarse categorical habitat thresholds which can be applied uniformly across the resiliency units (Table 7). Table 7 details our summary of current condition for the 15 resiliency units that we had enough information to analyze. Table 7: Categorical definitions used to assess the current condition of the resiliency units of Puget Oregonian. Temporal relevance of Number of Potential adjacent habitat Score occurrence occurrences Habitat Quality connectivity and quality High connectivity: Within High suitability: resiliency unit and in a one Assumed forest stand kilometer radius habitat with bigleaf maple tree Common: contains contiguous forest Contemporary: and understory habitat > 20 stand with bigleaf maple tree after 1994 (NWFP components at 75 to 3 observations and understory habitat Survey and 100 percent modeled in a resiliency components at 75 to 100 management) suitable habitat and on unit percent modeled suitable average greater than .6 habitat and on average greater value of NDVI values than .6 value of NDVI values for resiliency unit for resiliency unit. Suitable: Assumed forest stand with Moderate connectivity: bigleaf maple tree within resiliency unit and a component and one kilometer radius habitat Uncommon: understory habitat contains contiguous forest 20 - 10 components and 50 to stand with bigleaf maple tree Late century: 2 observations 75 percent modeled and understory habitat 1970 - 1994 in a resiliency suitable habitat and components at 50 to 75 unit average greater than .6 percent modeled suitable value of NDVI values habitat and on average greater for resiliency unit of than .6 value of NDVI values NDVI values for for resiliency unit. resiliency unit.

32 Low suitability: Low connectivity: within Assumed patches of resiliency unit and a one forest stand with kilometer radius habitat bigleaf maple tree contains contiguous forest Rare: component and stand with bigleaf maple tree Mid-century: 1 10 - 3 understory habitat and understory habitat 1950 - 1970 observations components and 25 to components at 25 to 50 50 percent modeled percent modeled suitable suitable habitat and a habitat and on a 0.4 to 0.6 0.4 to 0.6 NDVI value NDVI value for resiliency for resiliency unit. unit. Not suitable: no longer Isolated: Within a one habitat (land kilometer radius no habitat conversion, fire, exists (i.e. land conversion, Not viable or logging, no longer fire, logging, no longer maple Historic: earlier unknown: maple component) or 0 component) or less than 25 than 1950 2 or less less than 25 percent percent modeled suitable observations modeled suitable habitat and less than 0.4 value habitat and less than 0.4 of NDVI values for resiliency value of NDVI values unit. for resiliency unit.

When summarizing the resiliency level we averaged the score of each category for each unit. The average score was then used as an index of resiliency. We considered units with a score of zero for categories 1 and 2 to have an unknown resiliency status regardless of the average score. This was done because, based on what we know at this time, there appears to be areas of suitable habitat unoccupied by the snail and because we do not feel that we can assess resiliency for the unit or the species from a half-century old observation or a single observation of the species. Individual scores and resiliency unit ranks can be found in Appendix III.

33 Table 8: Known Current Condition of the resiliency units of the Puget Oregonian. Potential Temporal surrounding Number of Resiliency Resiliency Unit Name relevance of Habitat Quality habitat occurrences level* occurrence connectivity and quality Middle Columbia-Hood-4 Contemporary Rare High Suitability High Connectivity High Moderate Middle Columbia-Hood-7 Contemporary Rare High Suitability Moderate Connectivity Puget Sound-1 Contemporary Rare High Suitability High Connectivity High Puget Sound-6 Contemporary Rare Not Suitable Isolated Low Moderate Snoqualmie-1 Contemporary Rare High Suitability Moderate Connectivity Moderate Snoqualmie-2 Mid-Century Rare Suitable Low Connectivity Moderate Upper Cowlitz-2 Contemporary Common High Suitability High Connectivity Upper Cowlitz-3 Contemporary Common High Suitability High Connectivity High Upper Cowlitz-5 Contemporary Rare Low Suitability Low Connectivity Low Upper Cowlitz-6 Contemporary Common High Suitability High Connectivity High Upper Cowlitz-7 Contemporary Common High Suitability High Connectivity High Upper Cowlitz-11 Contemporary Rare High Suitability High Connectivity High Upper Cowlitz-13 Contemporary Rare High Suitability High Connectivity High Upper Cowlitz-15 Contemporary Rare High Suitability High Connectivity High Upper Cowlitz-16 Contemporary Rare High Suitability High Connectivity High * There are 59 additional identified and mapped resiliency units for this species which the Service was unable to determine a current resiliency condition at this time (Appendix III).

As indicated above, the majority of the resiliency units we analyzed had a high level (10) and only a few had moderate (2) or low (3) levels. This is more a result of sparse level of information about the species leading us to assume that where there are contemporary records of more than two observations per resiliency unit we can assume that there are enough individuals to constitute a recent “population” that must have some level of resiliency. It is important to recall that we are unable to determine a level of resiliency for the overwhelming (59 out of 74 or 80 percent) number of resiliency units.

Representation Table 9: Representative areas for the snail. The table also includes total number of resiliency units in each representative area and the level of resiliency for those units. For 19 of 23 units, or 82 percent, we have no understanding of the resiliency unit level within the representative area. State Representative Area Number of Resiliency Units within Current condition 7 Middle Columbia-Hood-4 High Middle Columbia-Hood Middle Columbia-Hood-7 High Middle Columbia-Hood-1, 2, 3, 5, 6 Unknown 7

WASHINGTON Puget Sound Puget Sound-1 High Puget Sound-6 Low

34 Puget Sound-2, 3, 4, 5, 7 Unknown 2 Snoqualmie Snoqualmie-1 High Snoqualmie-2 Moderate 17 Upper Cowlitz-2 High Upper Cowlitz-3 High Upper Cowlitz-5 Low Upper Cowlitz-6 High Upper Cowlitz Upper Cowlitz-7 High Upper Cowlitz-11 High Upper Cowlitz-13 High Upper Cowlitz-15 High Upper Cowlitz-16 High Upper Cowlitz-1, 4, 8, 9, 10, 12, 17 Unknown Deschutes 1 Unknown Duwamish 2 Unknown Hood Canal 1 Unknown Lake Washington 3 Unknown Lewis 6 Unknown Lower Cowlitz 3 Unknown Naches 1 Unknown Nisqually 4 Unknown Puyallup 3 Unknown Sauk 1 Unknown Stillaguamish 1 Unknown Upper Chehalis 1 Unknown Upper Yakima 1 Unknown Clackamas 2 Unknown Lower Columbia- 2 Unknown

Clatskanie

Lower Columbia-Sandy 1 Unknown Lower Willamette 3 Unknown OREGON Tualatin 1 Unknown Yamhill 4 Unknown

Redundancy Redundancy refers to the number of populations of a species and their distribution across the landscape, reflecting the ability of a species to survive catastrophic events. The greater the number of 35 populations/subpopulations, and the more widely they are distributed, the lower the likelihood a single catastrophic event will cause a species to become extinct. The Puget Oregonian is a moderately wide ranging PNW endemic species with assumed extant distribution across nearly 400 km (250 miles) north to south at elevations from sea level to 2700 feet 823 m) in mature hardwood forest habitat. The Service has identified 74 unique resiliency units, albeit unequally distributed across its range. We have enough information to analyze the condition of 15 of the units. Twelve of the 15 had “high’ resiliency, and there was at least one “high” resiliency unit in each of the four representative areas containing resiliency units that could be analyzed. We recognize that Puget Oregonian have a large current range and multiple representative areas. Because of this, catastrophic events are unlikely to cause the species to become extinct. However we were unable to evaluate the resiliency of the majority (80 percent) of individual “populations” across the range. Therefore at this time, redundancy is characterized by 74 populations with mostly undeterminable resiliency, well distributed across the range of the species. Compared to known historical condition, the extirpation of the populations in British Columbia and the eastern central Puget Sound populations indicate some level of decline in representation and redundancy over time.

Summary of Current Condition

The current condition of Puget Oregonian is characterized by 10 highly resilient populations, two moderately resilient population, three populations with low resiliency, and 59 populations with unknown resiliency (Figure 12). Redundancy appears adequate given these populations are distributed throughout the range of the species. Furthermore, the 74 populations or resiliency units are distributed across 23 representative areas throughout the range of the species, and the 10 highly resilient populations are distributed across four different representative areas.

36 State Boundaries Miles-======--====---- 0 25 50 100 Resiliency Unit Condition high moderate low unknown

Figure 11: Graphic depiction of Puget Oregonian resiliency units. Note this figure not only displays the limited knowledge of occupied locations for the species, but also the high level of uncertainty regarding the species resiliency across the species known and modeled distribution. The majority of the records and information about the species comes from the Cispus Watershed of Cowlitz County, Washington represented by the cluster of “high” resiliency units.` 37 FUTURE CONDITIONS

In this section, we forecast future conditions for the 15 sites for which we were able to assess the condition of Puget Oregonian. While there are multiple potential risk factors for the Puget Oregonian, we chose to focus on the variables we felt could most accurately forecast based on the data and models available and our knowledge of management at each of the sites. These variables include both the risk factors that we have described in previous sections (forest management, land use conversion, wildfire, bigleaf maple die-off), and climate change projections for temperature, precipitation, wildfire risk, and flood risk. We include four potential scenarios based on how the previously described risk factors and climate change might affect the sites. This analysis will help inform the potential risk of extinction for the Puget Oregonian at each site in the future. Variation in global temperatures is influenced by changes in solar radiation, volcanic eruptions, and greenhouse gas emissions (Mote et al. 2013, p.25). The extent and speed at which both global and regional climate will change depends on both the amount of future carbon emissions and how climate changes in response to those emissions. Projections of future climate conditions are developed by making assumptions about future greenhouse gas emissions and then modeling how climate changes in response to those emissions. Because there is uncertainty about both future emissions scenarios and how climate will respond, projections of future climate always include a range of scenarios. Climate models have constructive utility because they allow us to make predictions of how climate may change in the future, but their results should be interpreted cautiously. Models are mathematical representations of what can happen, but they do not always accurately predict future events. For our analysis of the Puget Oregonian’s future condition, we acknowledge the innate uncertainty associated with climate modeling as well as uncertainty with respect to how the species will respond to the effects of climate change. We also recognize that these models represent some of the best available scientific data we can utilize for predicting the species’ future condition. Climate Change Greenhouse gas emissions have increased at an unprecedented rate during the 20th century, resulting in global climate change (Intergovernmental Panel on Climate Change (IPCC) 2014, pp. 2-3). Climate change is likely to impact the Puget Oregonian through direct effects to individuals and populations, and indirect effects to suitable habitat. In order to incorporate these projected effects of climate change into future conditions for the Puget Oregonian, we analyzed predictions of two IPCC greenhouse gas emissions scenarios. The IPCC identifies various greenhouse gas Representative Concentration Pathways (RCPs) which take into account different scenarios of greenhouse gas emissions, atmospheric concentrations, and land use likely to unfold in the 21st century. The IPCC characterizes several potential scenarios, including RCP 4.5, an intermediate emissions scenario where atmospheric CO2 concentrations are expected to equal approximately 650 ppm after the year 2100, and RCP 8.5, where emissions aggressively increase to approximately 1370 ppm CO2 after the year 2100. For comparison, current atmospheric CO2 concentrations are around 400 ppm (IPCC 2014, p. 57). For the purposes of analyzing future conditions for the Puget Oregonian, we considered one intermediate scenario that assumes moderate cuts are made to emissions (RCP 4.5), and one high emissions scenario that assumes no deviation from the current emissions trajectory (RCP 8.5). These emissions scenarios were chosen because they frame the most likely high and low boundaries of the possible future in regard to greenhouse gas emissions. We consider future scenarios using the time period from the present to the middle of the century (2040-2069) (approximately 20-50 years), which is within the timeframe most climate models make projections for, as the uncertainty of future climate response to global warming increases with time from the present (IPCC 2014, p. 59). We analyzed the effects of climate change in areas that overlap with known Puget Oregonian populations through the middle of the century using data obtained from the Northwest Climate Toolbox, developed by

38 members of the Applied Climate Science Lab at the University of Idaho (Hegewisch et al. 2019). In addition to past and current data, the Northwest Climate Toolbox provides modeled future projections of climate, hydrology, and fire danger, based on the effects of potential degrees of greenhouse gas emissions reported by the IPCC (IPCC 2014, entire). Each future projection dataset we used for the purpose of analysis was a multi- model mean derived from multiple downscaled Coupled Model Intercomparison Project 5 (CMIP5) models. Though projections from individual models will vary for many reasons, the multi-model means often provide a good central estimate of the projected change (Hegewisch et al. 2019). Data and projections obtained from the Northwest Climate Toolbox provide estimates of future conditions, but may not be entirely accurate for any given site or year. Projected changes within the range of the Puget Oregonian that could affect the species include changes to temperature, precipitation, fire risk, and flood risk. Following global trends, average annual temperatures are expected to rise in the Pacific Northwest. At the same time, temperature variability is also expected to increase, leading to an expected increase in the number of warm days (warmer than 86 degrees Fahrenheit) during summertime. Overall, total annual precipitation will likely remain relatively constant with climate change, but variability in precipitation patterns is predicted to increase. Changes in precipitation vary by season, and location within the range of the snail. Throughout the entire range, summer precipitation is predicted to decrease, while winter precipitation is predicted to increase. In the northern portion of the snail’s range, spring and fall precipitation is expected to increase, while in the southern portion of the range, it is expected to decrease. The hotter and drier summer conditions will likely lead to higher fire risk, due to drier soils, and drier and more abundant fuels, created by the more droughty conditions. As part of the expected increase in winter precipitation, and the increased amount of precipitation expected to fall as rain rather than snow, we also expect an increase in the number and intensity of extreme precipitation events, which will increase the frequency and severity of floods. All of these changes are predicted to occur by the middle of the century (2040-206) under both RCP 4.5 and 8.5 scenarios, though at higher severity under the higher emissions scenario. For a more detailed description of climate change effects, refer to Appendix IV. The anticipated effects of climate change on the Puget Oregonian, and its interactions with existing risk factors are discussed in the Analysis of Future Conditions section. It is important for us to acknowledge uncertainty in the potential effects of climate change in our analysis of the Puget Oregonian’s future condition. While we expect climate change will negatively impact the Puget Oregonian, it is unclear based on our present knowledge of the species how these would affect the snail and how we would predict the impact of these effects on the populations into the future, as the specific environmental conditions that support Puget Oregonian in its habitat are unknown. However, we understand basic habitat and microhabitat conditions required by the Puget Oregonian and terrestrial mollusks in general to carry out their life cycles, which can inform our understanding of the influence that climate change may have on the Puget Oregonian. As discussed in the Habitat section, the species is associated with cool, moist conditions created by high canopy cover, and/or naturally wet areas created by springs or seeps. It also requires at least a partial component of bigleaf maple in the canopy, and ground-level shelter from leaf litter, swordfern skirts, and/or large woody debris. The effects of climate change discussed above, including changes in temperature, precipitation, soil moisture, flood risk, and fire risk, are all likely to have an effect on these habitat requirements and the Puget Oregonian in a way that may decrease the resilience of the species as a whole. Overall, we expect climate change to negatively impact the Puget Oregonian by decreasing suitability of current habitat on the landscape, decreasing or shifting the geographic area and seasonal/daily timing periods that provide suitable conditions for activity, and increasing the likelihood of events that could directly kill the species or destroy suitable habitat. However, the specific scope and magnitude of these potential effects are uncertain. Analysis of Future Conditions

39 Risk Factors As noted earlier, our Future Condition analysis focuses on the influence factors that we can forecast into the foreseeable future, including timber harvest, land use conversion, habitat loss, drought, wildfire risk, and flood risk. For the purpose of analyzing future conditions, we considered how those specific risk factors might combine to affect suitable habitat for the species, as we lack the information to determine how specific levels of change in the risk factors will impact the species. For example, there is inadequate data available to determine the demographic trends and requirements of the Puget Oregonian presently, let alone how the demographics might change into the future with impacts from climate change, land development, population growth, and other human related activities. Regarding the effects of climate change, though it appears that Puget Oregonian could have annual average precipitation and temperature limitations, it is difficult to predict what the effects of climate change on those limitations will mean for the species. For instance, while we can estimate how much summer precipitation will decrease in different portions of the snail’s range, we do not know if there is a threshold or critical value of temperature increase that result in certain negative effects to the species. Likewise, we know that summer temperatures will likely increase in the future, but we do not know what level of increase will prove too limiting for the snail’s activity and survival needs. However, we know generally what kinds of habitat and microhabitat conditions the Puget Oregonian requires for its different life stages and how different aspects of climate change and human factors may affect those conditions. In general, we expect the amount, quality, and connectivity of Puget Oregonian habitat to decline in the future, due to the effects of climate change and human activity, due to their expected effects on bigleaf maple, microhabitat conditions, habitat fragmentation, and more. We describe those expectations for future risk factors below. For the construction and analysis of future scenarios for the species, we will consider how those risk factors will impact habitat quality and connectivity, which were both also used to assess the current condition of each resiliency unit. Puget Oregonian is highly dependent on suitable habitat and microhabitat conditions, and therefore we assume that the current and future conditions of suitable habitat at the sites can approximate the current and future conditions of the species at those sites. As described in previous sections, there are many risk factors that can potentially affect the quality and connectivity of habitat. While all of these factors differ in how they might specifically affect suitable habitat for the Puget Oregonian, we consider the most important aspect of all these risk factors to be the fact that they lead to a loss in suitable habitat conditions, and therefore habitat quality and connectivity. More than one of the risk factors may affect any one location, but the occurrence of any one of them is likely sufficient to cause declines in habitat for the species. The occurrence of more than one risk factor at a location will likely lead to more severe habitat effects due to synergistic interactions; however, we do not have the means to quantify different levels of severity and their effects to the Puget Oregonian in a biologically meaningful way. Timber Harvest Most of the Puget Oregonian occurrence records, especially those observed in more recent years, are located on USFS lands, due to the species’ status as a Survey and Manage under the NWFP. As a result, locations where we have the highest amount of confidence about the presence and abundance for the snail are in locations where active vegetation management and timber harvest are occurring, and will likely continue to occur. Timber harvest results in loss of canopy cover and ground-level disturbance, leading to the loss of forest litter, ground- level vegetation, and large woody debris. All of these effects negatively impact habitat suitability for Puget Oregonian by removing food and shelter sources and creating hotter and drier microclimate conditions at the ground level. As multi-use management is one of the U.S. Forest Service’s core missions, timber harvest is likely to continue occurring on lands managed by the agency, likely leading to further habitat degradation in the future. However, we expect the extent and intensity of habitat degradation due to timber harvest will be relatively limited, due to surveys and conservation measures in place for Survey and Manage species under the NWFP. Timber. Other restrictions under the Northwest Forest Plan, including timber harvest restrictions in

40 stands older than 80 years, riparian reserves, and Late-Successional Reserves will also likely limit the effects of timber harvest. Timber harvest can also potentially help alleviate the effects of other risk factors that may negatively affect suitable habitat, such as wildfire and forest pests and diseases, by lowering tree densities and thereby hampering the spread of both fire and pests across the landscape. Additionally, while timber management does at times occur within mixed stands, stands composed mostly of bigleaf maples are not usually targeted for timber harvest, limiting potential habitat degradation, and potentially leading to long-term habitat improvement through bigleaf maple release. Therefore, we expect that significant amounts of Puget Oregonian habitat will remain unharvested into the future. Land Use Conversion and Urbanization As described in our discussion of the species range and habitat requirements, the Puget Oregonian is heavily reliant on forest habitats for survival and reproduction, due to the microhabitat conditions created by these environments. As described in previous sections, we expect the human population in the Pacific Northwest to increase in the future. As a result of this projected human population increase, urban development and sprawl outside federal lands are likely to increasingly encroach on forested lands that may serve as suitable habitat for the species. While some of the species records were recorded in locations that have been urbanized, most of these records are historic and likely no longer represent existing populations. As a result, we expect that future human population increases and the resulting urbanization and land use conversion will likely lead to a reduction of available species habitat. Loss of Bigleaf Maple and Suitable Habitat Bigleaf Maple is an important component of Puget Oregonian habitat, providing food and nutrients, and suitable microhabitat conditions for the snail. Like the snail, the bigleaf maple also thrives in moist soil conditions, which are likely to be negatively impacted due to projected changes in summer temperature and precipitation patterns. As a result, we expect the range of the bigleaf maple to either shift or contract to cooler and wetter locations. While range shift and contraction of tree species usually occurs slowly over time, the Puget Oregonian has very limited dispersal capabilities, and may not be able to adapt to the changes as quickly as bigleaf maple. Additionally, bigleaf maples are already declining due to maple dieback disease. The extent and severity of this disease are not fully known, but bigleaf maple tree mortality, dieback of branches and whole crowns has been observed in the species starting in 2009 (Hudec et al. 2019, p. 144-145). Climate change is a likely contributor to the spread of the disease, and will likely continue to negatively affect the amount and health of available bigleaf maple habitat. The severity of this decline in bigleaf maple habitat due to the disease is not fully known, and could thus could result in a large range of effect severity to the Puget Oregonian and its habitat, from isolated pockets of habitat decline, to extensive range-wide loss of suitable habitat for the species. Hotter and Drier Summer Conditions Due to longer, hotter, and drier summer conditions, Puget Oregonian activity will likely be more limited during the summer. Combined with more extreme projected climate conditions during winters, the snail will likely cease activity earlier and become active again later, both seasonally and daily. Due to the more extreme conditions expected due to climate change, periods during which the species can be active will likely also be shorter. These conditions will likely be limiting for the Puget Oregonian, as it limits the time during which the species can carry out the activities it requires for survival, reproduction, and dispersal. Additionally, it is unknown how long the species is able to remain in a state of estivation. If the increasing length of summer conditions extends beyond the amount of time the snail can remain alive in estivation or the species is unable to adjust the timing of their active periods, predicted climate change could lead to reduced survival of the species during summers. Alternatively, if the increasing length of summer conditions does not extend beyond the amount of time the snail can remain alive in estivation and the species is able to adjust the timing of their active periods, projected climate change may not affect the survival of the species during future summers. Increased Fire Risk 41 High-intensity forest fires likely kill Puget Oregonians directly, even when they are hidden underground, by creating excessively hot and dry conditions the species is unable to survive. Even low-intensity fires are likely to have negative effects on the species even if they do not directly kill the species, by removing leaf litter, killing understory vegetation, and potentially destroying downed logs and large woody debris, all of which the species relies on to create suitable microhabitat conditions on the ground. Forest fires are stochastic events, and it is thus difficult to predict if any one location will be affected by fires. However, we are reasonably certain that fire risk and fire intensity will increase across the range of the Puget Oregonian in the future, as climate change creates increasingly fire-prone landscapes, due to changes in summer temperature, precipitation, and fuel moisture. With the probability of fire occurrence and expected fire intensity increasing in the future across the snail’s range due to climate change, the risk of high-intensity fires destroying large contiguous areas of habitat, potentially removing entire populations of Puget Oregonian in the process, increases as well. Wetter Winter Conditions and Increased Flood Risk As described in the discussion of habitat for the Puget Oregonian, we suspect that the species is not tolerant to floods, because it is not often found in flood-prone areas, such as river floodplains, despite requiring high amounts of moisture for survival. With projected changes to winter precipitation due to climate change, we expect the frequency of flooding to increase. Additionally, due to expected increases in severity of extreme precipitation events, we expect larger areas to flood, including locations that are currently not flood-prone, where the Puget Oregonian might be found. As a result, we expect the amount of Puget Oregonian mortality to increase due to flooding in the future.

Future Scenarios For our analysis of the Puget Oregonian’s future condition at each site, we constructed four future scenarios focused on possible future trends in the quality and connectivity of habitat for the species (Table 10). These scenarios are meant to cover a large breadth of future conditions that could occur in Puget Oregonian populations, and all scenarios may not be equally plausible. To analyze future condition, we projected each scenario to a future time period, middle of the century, corresponding to climate modeling data. We expect climate change to negatively affect the macro- and microhabitat conditions required by the Puget Oregonian, and that the severity of change will be greater under a higher emissions scenario (RCP 8.5) than a lower emissions scenario (RCP 4.5). However, because we do not have enough information on the biology of the Puget Oregonian to determine how much habitat change due to climate change the species can tolerate and what threshold levels of change will result in significant effects on the species, it is at this time impossible for us to quantitatively differentiate between the effects of climate change under a low emissions scenario (RCP 4.5) and a high emissions scenario (RCP 8.5). Therefore, considering the significant magnitude of change occurring under even a lower emissions scenario (RCP 4.5), we assume that both emissions scenarios will result in similar negative effects on the species and its habitat, at different unquantifiable levels of severity out to mid- century (2040-2069).

Table 10: The four future scenarios used to estimate future conditions at each site.

Scenario #1 Scenario #2 Scenario #3 Scenario #4 No significant No significant Significant reduction Significant reduction reduction in habitat reduction in habitat in habitat quality in habitat quality quality quality No significant Significant reduction No significant Significant reduction reduction in habitat in habitat reduction in habitat in habitat connectivity connectivity connectivity connectivity Scenario 1 42 The first scenario should be considered a best-case scenario for the future conditions of the Puget Oregonian, in which both the habitat quality and connectivity remain near current conditions at a site. In order for this scenario to occur, the effects of climate change would have to remain mild, with temperature and precipitation conditions that are similar to current conditions, or within ranges that the snail and its habitat are able to tolerate. Such conditions might be possible if the site is located within a climate refugium, if currently present habitat is particularly robust and can persist (at least temporarily) despite climate change, or if the severity of climate change is significantly lower than expected (i.e. significantly below expectations under RCP 4.5). Human development and timber harvest will not have affected the site or its surroundings, in order to keep microhabitat and habitat conditions suitable and connected. Additionally, the site and its surroundings will have to have avoided catastrophic events, such as forest fires and flooding that could potentially destroy large patches of habitat. We consider this future scenario to be the least likely to occur, given current climate change predictions under even a lower emissions scenario (RCP 4.5) and factors such as maple dieback disease already affecting robustness of suitable habitat across the landscape. Habitat conditions in and around every site will likely face at least some decline due to expected changes in the species’ risk factors. Scenario 2 In the second scenario, no significant reduction in habitat quality would occur at a site, but suitable habitat would have faced significant declines around it. As in scenario 1, habitat quality at a site may remain constant if it is located within a climate refugium, if it is currently particularly robust, or if the severity of climate change effects is significantly lower than expected. However, in this scenario we expect that habitat connectivity will be negatively affected, possibly as a result of climate change effects, timber harvest, human encroachment, fires, or due to general habitat decline and bigleaf maple die-off happening around but not within the site. In this scenario, the Puget Oregonian may be able to persist at a site while it continues to retain its suitable habitat conditions; however, the species’ dispersal abilities may be greatly hindered due to the loss of connectivity, making it vulnerable to future habitat decline within the site. Scenario 3 The third scenario is similar to the second, in which one aspect of site condition remains constant while the other declines. However, the trends in the two site condition aspects is reversed: habitat quality within the site declines, while connectivity and habitat quality around the site remains constant. The reasons for which the condition of each aspect might decline or remain constant are similar to those described in scenario 2; they would just occur at different locations relative to the site. Under this scenario, we expect the condition of the species to decline within the site, due to the decline of suitable habitat conditions within the site. However, it may be possible for individuals of Puget Oregonian to disperse into surrounding suitable habitat if there is some available, especially for those individuals located near the edge of the site. Scenario 4 The fourth scenario would be the worst-case scenario in our analysis, in which both habitat quality and habitat connectivity decline. In this scenario, suitable habitat both within the site and surrounding the site would have undergone significant decline, likely due to the risk factors described above, including climate change, human encroachment, timber activity, and/or catastrophic events. Out of the four scenarios, the fourth scenario is most likely to lead to the complete loss of the Puget Oregonian at a site, especially if the severity of expected decline is high enough, and/or the current condition of the site is poor enough that the site becomes unsuitable for the species, and there are no opportunities for it to disperse elsewhere. Assumptions of Future Conditions We analyzed how the effects of the future scenarios changed the site condition rating for the locations with detections of the Puget Oregonian. We created three condition categories to characterize future site condition as described below. Similar to the current condition analysis, we arrived at an overall future condition at each site by looking at habitat quality and habitat connectivity. As discussed, we have no demographic data to determine 43 population trends of the species at each site. We thus assume that if a population is more robust today (it more common and recently detected), it has a better chance of persisting into the future. For habitat quality and connectivity, we consider the current habitat conditions and incorporate the four future scenarios to arrive at future habitat conditions. As we are unable to predict the severity of change in habitat conditions, we analyze a range of possible futures for each scenario, at each site. While we discuss the possibility of the presence of climate refugia that might serve to buffer the Puget Oregonian from the effects of climate change in some locations in the description of future scenarios, the scale of available climate projection data is at too coarse of a scale to determine where those refugia might be for the species and which specific sites might serve as climate refugia. Most available climate prediction datasets have a spatial resolution of about 820 ft (250 m), which is suitable for landscape-scale analysis and generalized conclusions about the effects of climate change. However, these data are too coarse-scale to differentiate between fine-scale future conditions to identify refugia at scales relevant to the species (i.e. a site down to about 33 ft (10 m). Therefore, for the purposes of our analysis of future scenarios at the sites (especially scenarios 2 and 3), we assume that any of the Puget Oregonian sites could either be a climate refugium, or be affected by the other risk factors in a way that will result in locally heterogeneous conditions. We characterized the future condition of the resiliency units according to the following overall condition categories:

• High: The site currently has relatively numerous and recent species records, and the site will retain relatively high habitat quality and connectivity, as defined in Table 7. Both the species and habitat are relatively likely to persist at the site.

• Moderate: Considering the four factors contributing to site conditions, as described above, overall conditions are intermediate, as defined in Table 7. Moderate conditions in the future may result from either moderate conditions of the four factors overall, or from some factors being high and others being low. For example, a site would be considered to be in moderate condition if species records are relatively numerous and recent, indicating some level of resilience, but habitat quality and connectivity are in low to moderate conditions. The species and/or habitat may persist at the site, but may not be robust to additional stressors.

• Low: The four factors contributing to site condition are in relatively poor condition individually, and overall, as defined in Table 7. Existing Puget Oregonian populations are likely not robust currently, and habitat conditions will be relatively poor. Both the species and habitat are unlikely to persist at the site. Detailed effects of future scenarios are outlined in Table 11 for sites with for which we had enough information to assess the potential resiliency of Puget Oregonian. Because we analyze a range of possible futures for each scenario (ranging from mild to severe declines in habitat suitability and/or connectivity, depending on the scenario), some sites may have a range of possible future site conditions under certain scenarios (e.g. low – moderate, moderate – high, low – high). The description for these ranges of site condition categories correspond to the three categories described above. For more details on the assessment of future conditions of each resiliency unit under each scenario, see Appendix III. Similar to our analysis of current conditions of the resiliency units, we were unable to analyze the future conditions of 59 resiliency units under the four future scenarios, due to insufficient data. The condition of those 59 units under the future scenarios remains unknown. Table 11. Conditions of 15 known Puget Oregonian resiliency units currently and under future scenarios. The remaining 59 units of unknown condition due to insufficient information have been omitted from this table. Current Condition Scenario 1 Scenario 2 Scenario 3 Scenario 4 Middle Columbia-Hood-4 Temporal relevance of Contemporary occurrence

44 Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Middle Columbia-Hood-7 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding habitat Moderate Moderate Isolated - Low Moderate Isolated - Low connectivity and Connectivity Connectivity Connectivity Connectivity Connectivity quality low - moderate low - moderate Site Condition moderate moderate moderate Puget Sound-1 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Puget Sound-6 Temporal relevance of Contemporary occurrence Number of Rare occurrences Habitat Quality Not Suitable Not Suitable Not Suitable Not Suitable Not Suitable Surrounding habitat Isolated Isolated Isolated Isolated Isolated connectivity and quality Site Condition low low low low low Snoqualmie-1 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat Moderate Moderate Moderate Moderate Moderate connectivity and Connectivity Connectivity Connectivity Connectivity Connectivity quality low - moderate low - moderate Site Condition moderate moderate moderate Snoqualmie-2 Temporal relevance of Mid-Century occurrence

45 Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality Suitable Suitable Suitable Low Suitability Low Suitability Surrounding habitat Moderate Moderate Isolated - Low Moderate Isolated - Low connectivity and Connectivity Connectivity Connectivity Connectivity Connectivity quality Site Condition low low low low low Upper Cowlitz-11 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Upper Cowlitz-13 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Upper Cowlitz-15 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Upper Cowlitz-16 Temporal relevance of Contemporary occurrence Number of Rare occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality low - moderate Site Condition high high moderate moderate Upper Cowlitz-2 Temporal relevance of Contemporary occurrence 46 Number of Common occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding habitat Moderate Moderate Isolated - Low Moderate Isolated - Low connectivity and Connectivity Connectivity Connectivity Connectivity Connectivity quality moderate - high moderate - high low - moderate Site Condition high high Upper Cowlitz-3 Temporal relevance of Contemporary occurrence Number of Common occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality moderate - high moderate - high low - high Site Condition high high Upper Cowlitz-5 Temporal relevance of Contemporary occurrence Number of Rare occurrences Habitat Quality Low Suitability Low Suitability Low Suitability Not Suitable Not Suitable Surrounding habitat Low Connectivity Low Connectivity Isolated Low Connectivity Isolated connectivity and quality Site Condition low low low low low Upper Cowlitz-6 Temporal relevance of Contemporary occurrence Number of Common occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality moderate - high moderate - high low - high Site Condition high high Upper Cowlitz-7 Temporal relevance of Contemporary occurrence Number of Common occurrences Not Suitable - Not Suitable - Habitat Quality High Suitability High Suitability High Suitability Suitable Suitable Surrounding Isolated - Isolated - habitat High Connectivity High Connectivity Moderate High Connectivity Moderate connectivity and Connectivity Connectivity quality moderate - high moderate - high low - high Site Condition high high

Summary of Future Condition

47 Overall, our future scenarios project that the future condition of Puget Oregonian will likely decrease across the range of the species. While we were only able to evaluate the current and future conditions of 15 out of the 74 resiliency units, we expect this general decline in conditions to occur within all resiliency units, even the 59 resiliency for which we did not have sufficient data for analysis.

Table 12. Summary of current and future condition of Puget Oregonian in future scenarios. The remaining 59 units of unknown condition due to insufficient information have been omitted from this table. Current Representative Area Scenario 1 Scenario 2 Scenario 3 Scenario 4 condition Middle Columbia-Hood low - Middle Columbia-Hood-4 high high moderate moderate moderate low - low - Middle Columbia-Hood-7 moderate moderate moderate moderate moderate Puget Sound low - Puget Sound-1 high high moderate moderate moderate Puget Sound-6 low low low low low Snoqualmie low - low - Snoqualmie-1 moderate moderate moderate moderate moderate Snoqualmie-2 low low low low low Upper Cowlitz moderate - moderate - low - Upper Cowlitz-2 high high high high moderate moderate - moderate - low - high Upper Cowlitz-3 high high high high Upper Cowlitz-5 low low low low low moderate - moderate - low - high Upper Cowlitz-6 high high high high moderate - moderate - low - high Upper Cowlitz-7 high high high high low - Upper Cowlitz-11 high high moderate moderate moderate low - Upper Cowlitz-13 high high moderate moderate moderate low - Upper Cowlitz-15 high high moderate moderate moderate low - Upper Cowlitz-16 high high moderate moderate moderate

Future Resiliency With the information we have, we anticipate the future resiliency of the 15 assessed resiliency units to decline in three out of the four scenarios by the middle of the century. We do expect resiliency to decline across the range of the species, but are unable to predict the severity of decline within each resiliency unit, as discussed in previous sections. While we cannot predict if there will be loss of any resiliency units, we anticipate that the lower level of resiliency overall will make the species more vulnerable to stochastic events. Future Representation

48 Representation for Puget Oregonian is difficult to project because we do not have enough information to project the probability of persistence of any particular resiliency unit. Future Redundancy The 15 resiliency units analyzed are found in 4 representative areas. The future condition of the resiliency units in these areas summarized in Table 12. At the current time we recognize 60 representative areas, but we are unable to determine their overall contribution to the redundancy of the species on the landscape. While we have general expectation for landscape-scale changes to habitat conditions in the future, we do not have enough information to determine if declining future conditions will result in a loss in representative areas.

Synthesis The SSA process for the Puget Oregonian was limited in its ability to assess the current viability of the species across its range. The Upper Cowlitz sub-basin contains the majority of observation and biological information on the species, making it the basis for most of our assumptions about the species’ viability. Based on available survey data from the Upper Cowlitz sub-basin, we understand the species to be fairly common and well distributed within suitable habitat. Outside of this sub-basin, species records are relatively rare, limiting our ability to analyze the resiliency, redundancy, and representation of the species across its range. However, we were generally able to identify habitat characteristics known to be important to the species. These habitat characteristics include bigleaf maple stands or mixed bigleaf maple stands, and microhabitat features such as large woody debris, swordfern, and leaf litter, that contribute to moist and cool conditions on the forest floor. To assess future conditions of the species out the middle of the century, we considered the effects of climate change on the species’ modeled habitat requirements. We also considered other risk factors, including human population growth in the Pacific Northwest, forest management, and disease. We identified four future scenarios, one which anticipates no change from current conditions, and three scenarios which anticipate various levels of decline in habitat conditions. Similar to the assessment of current condition, we did not have enough information to assess the future condition of 59 resiliency units for which we lack information. There is a wide range of possibilities for the future of the 15 resiliency units for which we did have sufficient information, as described in Table 12. We expect the resource needs of the species to diminish in quantity and quality as stressors to them increase in the future. Overall, we expect the viability of the species to decline under the future scenarios, but at the middle of the century we expect the species will still persist.

49 Literature Cited Abele, S.E. 2010. Gastropod diversity in the boreal mixedwood forest of northern Alberta-variation among forest types and response to partial harvesting. MS Thesis. Edmonton, Alberta: University of Alberta. 107 pp. Agee, J. K. 1996. Fire Ecology of Pacific Northwest Forests. Island press. 505 pp. Ansart A, P. Vernon, J. Daguzan J. 2001. Freezing tolerance versus freezing susceptibility in the land snail Helix aspersa (Gastropoda: Helicidae). Cryo Letters 22: 183–190. Arter, H.E. 1990. Spatial relationship and gene flow paths between populations of the alpine snail Arianta arbustorum (Pulmonata: Helicidae). Evolution. 44(4), 966-980. Baker, H.B. 1958. Land snail dispersal. The Nautilus. 71(4), 141-148 p. Barker, G.M. (Ed.). 2001. The Biology of Terrestrial Molluscs. CABI. 558 pp. Barnes, R. D. 1987. Invertebrate zoology 5th edition, Series. Saunders series in organismic biology. Saunders College Publishing. Philadelphia, PA. 893 pp. Baur, B. 1988. Microgeographical variation in shell size of the land snail Chondrina clienta. Biological Journal of the Linnean Society, 35(3), 247-259. Baur, B. 1994. Parental care in terrestrial gastropods. Experientia 50, 5–14 Betzen, J.J. 2018. Bigleaf Maple Decline in Western Washington. MS Thesis. University of Washington. Seattle, WA. 129 p. Bolsinger, C. L. 1973. Changes in commercial forest area in Oregon and Washington 1945-1970. Research Bulletion. PNW-RB-046. Portland, OR: US Department of Agriculture, Forest Service, Pacific Northwest Research Station. 16 p. Bouchet, P., J.P. Rocroi, J. Frýda, B. HausdorfW. Ponder A. Valdésand A. Warén. 2005. Classification and nomenclator of gastropod families. International Journal of Malacology. p. 410 Bouchet, P., Rocroi, J.P., Hausdorf, B., Kaim, A., Kano, Y., Nützel, A., Strong, E.E. 2017. Revised Classification, Nomenclator and Typification of Gastropod and Monoplacophoran Families. Malacologia, 61(1-2), 1-527. Bretz, J. H. 1913. Glaciation of the Puget Sound region. Washington Geological Survey Bulletin. 8(1) 244 p. British Columbia Invertebrates Recovery Team. 2008. Recovery strategy for Puget Oregonian Snail (Cryptomastix devia) in British Columbia. B.C. Ministry of Environment, Victoria, BC. 21 pp. Bureau of Land Management (BLM). 2019. Geographic Biotic Observations (GeoBOB). Bureau of Land Management. Portland, OR. Burke, T.E., Kogut, T., Duncan, N., Jordan, S. F., and Black, S. H. 1999. Conservation Assessment for Cryptomastix devia, Puget Oregonian. USDA Forest Service Region 6 and USDI Bureau of Land Management. Portland, OR. p. 24 Burke, T.E. 2013. Land Snails and Slugs of the Pacific Northwest. Oregon State University Press, Corvallis, OR. Committee on the Status of Endangered Wildlife in (COSEWIC). 2013. COSEWIC status appraisal summary on the Puget Oregonian Cryptomastix devia in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. xv pp.

50 Crisafulli, C.M., Clayton, D.R., and Olson, D.H. 2008. Conservation assessment for the Larch mountain salamander (Plethodon larselli). USDA Forest Service Region 6 and USDI Bureau of Land Management Interagency Special Status and Sensitive Species Program. 36 p. Curry, T.R., N. Greenwald, A. Garty. 2008. Petition to list 32 mollusk species from freshwater and terrestrial ecosystems of the northwestern United States as Threatened or Endangered under the Endangered Species Act. Center for Biological Diversity. Portland, OR. 85 pp. Dennison, P. E., Brewer S. C., Arnold J. D, Moritz M. A. 2014. Large wildﬁre trends in the western United States, 1984–2011, Geophysical Research Letters. 41 pp. 2928–2933 Duncan, N., T. Burke, S. Dowlan, P. Hohenlohe. 2003. Survey Protocol for Survey and Manage Terrestrial Mollusk Species Version 3.0. Bureau of Land Management, Roseburg, OR. Easterbrook, D. J., D. R. Crandell, E. B. Leopold. 1967. Pre-Olympia Pleistocene stratigraphy and chronology in the central Puget Lowland, Washington. Geological Society of America Bulletin, 78(1), 13-20. Edworthy, A.B., K.M.M. Steensma, H.M. Zandberg, P.L. Lilley. 2012. Dispersal, home-range size, and habitat use of an endangered land snail, the Oregon forestsnail (Allogona townsendiana). Canadian Journal of Zoology, 90(7), 875-884. Frest, T.J. and E.J Johannes. 1993. Mollusc Species of Special Concern Within the Range of the Northern Spotted Owl: With an Addendum Addressing New Management Options Proposed in June, 1993. Deixis Consultants. Frest, T.J. and E.J Johannes. 1995. Interior Columbia Basin mollusk species of special concern. Prepared for: Interior Columbia Basin Ecosystem Management Project. Seattle, WA: Deixis. Contract report. Frest, T.J. and E.J Johannes. 2000. A Baseline Mollusk Survey of Southwestern Oregon, with emphasis on the Rogue and Umpqua River Drainages. Final Report prepared for Oregon Natural Heritage Program, Portland, Oregon. Deixis Consultants, Seattle, Washington. vii + 403 pp., appendices. Frest, T.J. and E.J Johannes. 2005. Grazing Effects on Springsnails, Cascade-Siskiyou National Monument, Oregon. Report 2004. Final Report to World Wildlife Fund, Ashland, Oregon. Deixis Consultants, Seattle, Washington. iv + 182 pp Fried, J. S., Boyle, J. R., Tappeiner II, J. C., and Cromack Jr, K. 1990. Effects of bigleaf maple on soils in Douglas-fir forests. Canadian Journal of Forest Research, 20(3), 259-266. Foltz Jordan, S., S.H. Black. 2012. Effects of Forest Land Management on Terrestrial Mollusks: A Literature Review. The Xerces Society for Invertebrate Conservation, Portland, Oregon. p. 87 Foltz Jordan, S., S.H. Black. 2015. Conservation Assessment for Cryptomastix devia, Puget Oregonian. USDA Forest Service Region 6 and USDI Bureau of Land Management. Portland, OR. P. 32 Forsyth, R. G. 2004. Land Snails of British Columbia. Royal BC Museum. p. 188 Foster, A. D., J. Ziegltrum. 2013. Riparian-associated gastropods in Western Washington: Community composition and the effects of forest management. Northwest Science, 87(3), 243-256. Furnish, J., T. Burke, T. Weasma, J. Applegarth, N. Duncan, R. Monthey, D. Gowan. 1997. Survey Protocol for Terrestrial Mollusk Species from the Northwest Forest Plan Draft Version 2.0. Bureau of Land Management, Salem, OR. Gallant, J.C., J.P. Wilson. 2000. Primary topographic attributes. Terrain Analysis: Principles and Applications. Wiley, New York: 51–85.

51 Ganguly, A., K. Steinhaeuser, D. Erickson, M. Branstetter, E. Parish, N. Singh, J. Drake, L. Buja. 2009. Higher trends but larger uncertainty and geographic variability in 21st century temperature and heat waves. PNAS. 106: 15555–15559. Global Biodiversity Information Facility (GBIF). 2019. Cryptomastix devia occurrence information [raw dataset]. Gould, A.A. 1846. Shells collected by the United States Exploring Expedition under the command of Charles Wilkes. In Proceedings of the Boston Society of Natural History. Vol. 2, No. 3. Gould, S.J., Young, N.D., & Kasson, B. 1985. The consequences of being different: sinistral coiling in Cerion. Evolution, 39(6), 1364-1379 Haggard, J.A. 2000. A radio telemetric study of the movement patterns of adult northern red-legged frogs (Rana aurora aurora) at Freshwater Lagoon, Humboldt County, California MAB thesis. Arcata, CA: Humboldt State University. Harmon, M.E., J. F. Franklin, F.J. Swanson, P. Solllns, S.V. Gregory, J.O. Lattin, N.H. Anderson, S.P. Cline, N.G. Aumen, J.R. Sedell, G.W. Lienkaemper, K. Cromack, jr., K.W. Cummins. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15: 133-302. Hausdorf, B., and Hennig, C. 2003. Nestedness of North-West European Land Snail Ranges as a Consequence of Differential Immigration from Pleistocene Glacial Refuges. Oecologia. 135(1), 102-109. Hegewisch, K.C., J.T. Abatzoglou, O. Chedwiggen, B. Nijssen. Climate Mapper web tool. NW Climate Toolbox. Online Linkage: . (Accessed April 2019) Heller, J. 1990. Longevity in Molluscs. Malacologia, 31(2), 259-295. Hudec, J.L., J.E. Halofsky, D.L. Peterson, J.J. Ho. 2019. Climate Change Vulnerability and Adaptation in Southwest Washington. U.S. Department of Agriculture. Portland, OR. xvii + 252 pp. Hyman, L. H. 1967. The Invertebrates: Mollusca I (Vol. 6). McGraw-Hill. p. 792 iNaturalist. 2019. Cryptomastix devia occurrence information [raw dataset]. Integrated Digitized Biocollections (iDigBio). 2019. Cryptomastix devia occurrence information [raw dataset]. International Panel on Climate Change (IPCC). 2014. Annex II: Glossary. Mach, K.J., S. Planton and C. von Stechow (Eds.). In: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Core Writing Team, R.K. Pachauri and L.A. Meyer (Eds.). IPCC, Geneva, Switzerland, pp. 117-130. Johannes E.J. 2012 Events Leading to the Proposed Listing of 26 U.S. Snails and Slugs as a Result Of President Clinton’s Northwest Forest Plan. Tentacle Vol. 20 p. 7 - 9 Johannes E.J. 2017a. Correlating personal weather station data with Allogona townsendiana activity observations at Crystal Springs Park, Tukwila, Washington. The Dredgings, Newsletter of The Pacific Northwest Shell Club, Vol 57. No. 2, pp 3-5. Johannes E.J. 2017b. Correlating Personal Weather Station Data with Observations of the Activity of Allogona Townsendiana at Crystal Springs Park, Tukwila, Washington State, USA. Tentacle The Newsletter of the IUCN/SSC Mollusc Specialist Group 25 pp. 25-27. Jones, C.G., Lawton, J. H., Shachak, M. 1994. Organisms as ecosystem engineers. Ecosystem Management. Springer, New York, NY. pp. 130-147

52 Kelley, R., S. Dowlan, N. Duncan, T. Burke. 1999. Field Guide to Survey and Manage Terrestrial Mollusk Species from the Northwest Forest Plan. Bureau of Land Management, Oregon State Office. Portland, OR. Kozloff, E.N. 1976. Plants and animals of the Pacific Northwest: An illustrated guide to the natural history of western Oregon, Washington, and British Columbia. Seattle: University of Washington Press. 264 pp. Lewis, M. 1939. Historical Sketches, Fire History, Columbia National Forest 1939. Unpublished Report. USDA Little, Jr. E.L. 1971. Atlas of United States Trees. Volume 1. U.S. Department of Agriculture, Washington, DC. MacLean, C. D., and C. L. Bolsinger. 1997. Urban expansion in the forests of the Puget Sound region. Research Bulletin. PNW-RB-225. Portland, OR: US Department of Agriculture, Forest Service, Pacific Northwest Research Station. 17 pp. Marcot, B. G., & Vander Heyden, M. 2001. Key ecological functions of wildlife species. DH Johnson and TA O'Neil, Technical Coordinators. Wildlife-habitat relationships in Oregon and Washington. Oregon State University Press, Corvallis OR, 168-186. Marco, B. 2011. The Ecology of Skirts. https://www.plexuseco.com/EPOW/EPOW- Archive/archive_2011/EPOW-110919.htm. Accessed April 16, 2019 9:02:41 AM Master, L.L., D. Faber-Langendoen, R. Bittman, G.A. Hammerson, B. Heidel, L. Ramsay, K. Snow, A. Teucher, and A. Tomaino. 2012. NatureServe Conservation Status Assessments: Factors for Evaluating Species and Ecosystem Risk. NatureServe, Arlington, VA. pp. 64 Meehl, G.A.; T. F Stocker, W. D. Collins, P. Friedlingstein, T. Gaye, J.M. Gregory, A. Kitoh, R. Knutti, J.M. Murphy, A. Noda, S. C. B. Raper, I. G. Watterson, A.J. Weaver, Z. C. Zhao. Global climate projections. In: Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K. B. Averyt, M. Tignor, H.L. Miller. Editors, editors. IPCC. 2007. Climate Change 2007: The physical science basis. contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press; 2007. http://hdl.handle.net/102.100.100/124551?index=1. Mote, P.W., JT. Abatzoglou, and K.E. Kunkel. 2013. Climate: variability and change in the past and the future: Understanding global and regional climate change. Chapter 2 in Dalton, M.M., P.W. Mote, and A.K. Snover [Eds.]. Climate Change in the Northwest: Implications for Our Landscapes, Waters, and Communities. Washington, DC: Island Press. xl + 230 pp. NatureServe 2002. Element Occurrence Data Standard in cooperation with the Network of Natural Heritage Programs and Conservation Data Centers 201 p. http://downloads.natureserve.org/conservation_tools/element_occurence_data_standard.pdf NatureServe. 2019. Cryptomastix devia Puget Oregonian Snail webpage. http://explorer.natureserve.org/servlet/NatureServe?searchName=Cryptomastix+devia, accessed May 16, 2019 Nekola, J.C. 2014. Overview of the North American terrestrial gastropod fauna. American Malacological Bulletin, 32(2), 225-235. Omdal, D and Ramsey-Kroll, A. 2012. Assessing the role of Verticillium wilt in Bigleaf Maple (Acer macrophyllum) dieback in Western Washington. Washington State Department of Natural Resources. Olympia, WA. Poster publication. Omernik, J.M., and A.L. Gallant. 1986. Ecoregions of the Pacific Northwest (with map). EPA/600/3-86/033. U.S. Environmental Protection Agency, Corvallis, OR. 39 p. 53 Oregon Biodiversity Information Center (ORBIC). 2019. Invertebrate Animal Ranks. https://inr.oregonstate.edu/orbic/rare-species/ranking-documentation/invertebrate-animal-ranks. Ovaska, K.E., R.G. Forsyth. 2002. COSEWIC Assessment and Status Report on the Puget Oregonian Snail Cryptomastix devia in Canada. Committee on the Status of Endangered Wildlife in Canada. Ottawa. vi + 22 pp. Pechman, M. A. 2006. Case Number C04-844-P. U.S District Court, Western District of Washington at Seattle. Northwest Ecosystem Alliance et al. v. Mark R. Rey et al. Perez, K.E., Defreitas, N., Slapcinsky, J., Minton, R. L., Anderson, F. E., Pearce, T. A. 2014. Molecular phylogeny, evolution of shell shape, and DNA barcoding in Polygyridae (Gastropoda: Pulmonata), an endemic North American clade of land snails. American Malacological Bulletin, 32(1), 1-32. Peterson, E.B., Columbia, B., Columbia, B., Hardwood, B. C., and Vegetation Management Technical Advisory Committee. 1999. Bigleaf maple managers' handbook for British Columbia. BC Ministry of Forests, Forestry Division Services Branch, Production Resources. 107 pp. Phillips, S.J., M. Dudík, R.E. Schapire. 2017. Maxent software for modeling species niches and distributions (Version 3.4.1). Online Linkage: . (Accessed May 16, 2019). Pilsbry, H. A. 1939. The Land Mollusca of North America (North of Mexico), Vol. 1 part 1. Academy of Natural Sciences of Philadelphia, Monographs 3 I - 573573 pp. Pilsbry, H. A. 1940. The Land Mollusca of North America (North of Mexico), Vol. 1 part 2. Academy of Natural Sciences of Philadelphia, Monographs 3 i – ix, 575 - 994 pp. Pilsbry, H. A. 1948. The Land Mollusca of North America (North of Mexico), Vol. 1 and 2, parts 1 and 2 . Academy of Natural Sciences of Philadelphia, Monographs 3 i - xlvii, 521 - 1113 pp. Platt, John R. 2016. Snails Are Going Extinct: Here's Why That Matters. Extinction Countdown. Scientific American. https://blogs.scientificamerican.com/extinction-countdown/snails-going-extinct/. Accessed April 18, 2019. Pojar, J., MacKinnon, A., and Alaback, P. B. 1994. Plants of coastal British Columbia. Lone Pine Publishing. 528 pp. Prinn, R., S. Paltsev, A. Sokolov, M. Sarofim, J. Reilly, H. Jacoby. 2011. Scenarios with MIT integrated global systems model: significant global warming regardless of different approaches. Climatic Change, 104(3-4), 515-537. PRISM Climate Group. 2015. 30-yr Normal Precipitation: Annual, Period: 1981-2010. Oregon State University, Corvallis, OR. Online Linkage: . (Accessed May 14, 2019).\ Rees, W. J. 1965. The aerial dispersal of Mollusca. Journal of Molluscan Studies, 36(5), 269-282. Rouse, J.W., Jr., R.H. Haas, J.A. Schell, D.W. Deering. 1974. Monitoring vegetation systems in the Great Plains with ERTS. Proceedings, 3rd Earth Resource Technology Satellite (ERTS) Symposium 1. National Aeronautics and Space Administration, Washington, DC. Ruppert, E. E., Barnes, R. D., and Fox, R. S. 2004. Invertebrate Zoology: a Functional Evolutionary Approach. 7th Edition. Brooks/Cole-Thompson Learning, Belmont. CA. Schilthuizen, M., and Lombaerts, M. 1994. Population structure and levels of gene flow in the Mediterranean land snail Albinaria corrugata (Pulmonata: Clausiliidae). Evolution, 48(3), 577-586.

54 Schuett-Hames, J. 2004. Northern Red-Legged Frog (Rana aurora aurora) Terrestrial Habitat Use in the Puget Lowlands of Washington. MES Thesis. The Evergreen State College. Olympia, WA. 129 pp. Solem, A., and A.C. van Bruggen (Editors). 1984. World-wide snails: Biogeographical studies on non-marine Mollusca. Brill Archive. 289 pp. Steensma, K. M., P. L. Lilley, and H. M. Zandberg. 2009. Life history and habitat requirements of the Oregon forestsnail, Allogona townsendiana (Mollusca, Gastropoda, Pulmonata, Polygyridae), in a British Columbia population. Invertebrate Biology, 128(3), 232-242. Sudworth, G. B. 1967. Forest trees of the Pacific slope. US Government Printing Office. 455 pp. Szaro, R. C., Severson, K. E., and Patton, D. R. 1988. Management of amphibians, reptiles, and small mammals in North America: Proceedings of the symposium; 1988 July 19-21, Flagstaff, AZ. General Technical Report- RM-166, US Department of Agriculture, Forest Service, Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, 458 p. Tarrant, R.F., Isaac, L.A., and Chandler, R.F. 1951. Observations on litter fall and foliage nutrient content of some Pacific Northwest tree species. Journal of Forestry, 49(12), 914-5. Thomas, J.W. 1979. Wildlife habitats in managed forests: the Blue Mountains of Oregon and Washington. Agriculture Handbook No. 553. U.S. Department of Agriculture, Forest Service. 512 p. Tomiyama, K and M. Nakane. 1993. Dispersal patterns of the giant African snail, Achatina fulica (Ferussac) (Stylommatophora: Achatinidae). Equipped with a radio-transmitter. Journal of Molluscan Studies. 59- 315- 322 Turgeon, D.D., Quinn, J.F., Bogan, A.E., Coan, E.V., Hochberg, F G., Lyons, W. G., and Roth, B. 1998. Common and scientific names of aquatic invertebrates from the United States and Canada: Mollusks. American Fisheries Society, Special Publication 26. p. 536 Tyson, A. 2018. Timber Tracking - 2018 Report Adventure Scientists. Bozeman, MT. 38 p. https://www.adventurescientists.org/project-reports-and-scientific-publications.html United Stated Department of Agriculture (USDA) and United State Department of the Interior (USDI). 1994. Record of decision for amendments to Forest Service and Bureau of Land Management planning documents within the range of the northern spotted owl. USDA. 2011. Wildlife Biological Assessment (BA) for the North Fork Project; Cowlitz Valley Ranger District; Gifford Pinchot National Forest. U.S. Forest Service, Randle, WA. United State Fish and Wildlife Service (USFWS) 2019. https://www.fws.gov/endangered/esa-library/pdf/5- Year%20Listing%20Workplan%20May%20Version.pdf. Accessed 16-July-2019 09:47am United States Geologic Survey (USGS). 2006. Shuttle Radar Topography Mission, 1 Arc Second, Filled Finished 2.0, February 2000. Global Land Cover Facility, University of Maryland, College Park, MD. Online Linkage: . (Accessed May 14, 2019). USGS 2018. eMODIS Normalized Difference Vegetation Index Continental US, August 2018. Earth Resources Observation and Science (EROS) Center, United States Geologic Survey, Sioux Falls, SD. Online Linkage: . (Accessed May 14, 2019). USGS. 2019. Biodiversity Information Serving our Nation (BISON). Reston, VA. Vagvolgyi, J. 1968. Systematics and Evolution of the Genus Triodopsis (Mollusca: Pulmonata: Polygyridae). Volume 135, Number 7. Museum of Comparative Zoology, Harvard University. Cambridge, MA.

55 Washington, State of. 2018. Forecast of the State Population December 2018. Forecasting & Research Division, Office of Financial Management. Olympia, Washington. pp. 46 Waterstrat, F. T. 2019. Notes on rapid human assisted dispersal of the Pacific sideband snail in Thurston County. The Dredgings, Newsletter of The Pacific Northwest Shell Club, Vol 59. No. 5, pp 5-6. Weil R.R. and Brady N. C. 2002. Elements of the Nature and Properties of Soils 2nd Edition. Pearson Publishing. Upper Saddle River, NJ. 606 p. Wolf, S., Hartl, B., Carroll, C., Neel, M. C., & Greenwald, D. N. 2015. Beyond PVA: Why recovery under the Endangered Species Act is more than population viability. BioScience, 65(2), 200-207. Xerces Society for Invertebrate Conservation (Xerces). 2019. Puget Oregonian snail (Cryptomastix devia) spatial database. Accessed April 16, 2019. https://www.xerces.org/

Personal Communications

Donahey, James, Silviculturalist, United States Department of Agriculture, Forest Service. Email Correspondence with F. T. Waterstrat, July 30, 2019 Foster, Alex, Ecologist, US United States Department of Agriculture, Forest Service. Email Correspondence with Teal Waterstrat and Rebecca Migala, USFWS, November 4, 2019. Regarding Peer Review of the Draft Puget Oregonian snails Species Status Assessment. Kogut, T. Retired District Biologist. United States Department of Agriculture, Forest Service. Verbal Communication with A. M. Le and F. T. Waterstrat, USFWS. May 10 2019. Comments while on a field visit to learn about the Puget Oregonian Snail Le, A. M. and Waterstrat F. T. Fish and Wildlife Biologists. United States Department of Interior, Fish and Wildlife Service. Cispus Watershed, Cowlitz County, WA. May 10, 2019. Observations while on a field visit to learn about the Puget Oregonian Snail

56 Appendix I Molluscan Taxonomy, Physiology, and Ecology

Phylum Mollusca Phylum Mollusca is the second most diverse Phylum of animals on Earth exceeded only by Arthropoda (Hyman 1967, vol VI p. v., Barnes 1987, p.342, Ruppert et al. 2004 p. 284). The Phylum Mollusca contains a huge variety of life from the large, complex, and intelligent cephalopods to the minute and “simple microsnails.” Gastropoda – slugs and snails Of all molluscan classes Gastropoda is the largest, most diverse with more than 62,000 representative species and arguably successful both in form, habit, and habitat, occupying a multitude of marine, freshwater, and terrestrial habitats (The Mollusca: https://ucmp.berkeley.edu/taxa/inverts/mollusca/gastropoda.php accessed June 14, 2019, Barnes1987, p.342, 347). Gastropod or “stomach foot” is a reference to the anatomical position of the visceral mass above the foot of a specimen. In general gastropods have a well-developed head with one or typically two sets of sensory tentacles and a radula for feeding, possess a univalve spirally coiled shell (or without a valve in some slugs), a well-developed foot with a flat creeping sole, and a torted, or twisted body plan, independent of the valve (Hyman 1967, p. 152, Barnes 1987 347 -348) (Figure 1). This last characteristic, the ancestral 180 degree torsion in the visceral mass, is the unifying characteristic of gastropods and is still present to some degree in representatives of the taxon or their ancestors (Ruppert et al 2004, p. 303). The vast majority of gastropods are dextral (right) coiling species although sinstral (left) coiled forms are known to occur (Gould 1985, entire). Gastropoda includes the only terrestrial representatives of the Mollusca Phylum (Barker 2001, p 10). Former Subclass/Order/current informal group Pulmonata – slugs and snails with “lungs” or “land snails” This morphologically grouping of gastropods includes terrestrial and some freshwater and marine gastropod species. The primary characteristic that unifies this group is the loss of the gill and conversion of the mantle cavity into a lung or lung like structure and the absence of an operculum, a corneous or calcareous lid or door that covers a snail’s aperture (Ruppert et al 2004, p. 342, Burke 2013, p 10). The lung is what enables slugs and snails to colonize terrestrial habitats and breathe air (Figure 1). The roof (dorsal) portion of mantle is highly vascularized allowing for the passive exchange of gases through an opening in the mantle cavity called the pneumostome (Ruppert et al 2004, p. 342). Active gas exchange, or breathing, can be achieved by the Pulmonate gastropod by contracting and relaxing the mantle cavity causing flow of air through the pneumostome (Ruppert et al 2004, p. 342). Locomotion in Pulmonate, like that of gastropods, is conducted with the foot of the organism. Pulmonates use a large pedal, or mucus gland, to excrete a mucus trail over which the individual glides propelling itself with waves of muscular contractions along the sole of its foot (Ruppert et al 2004, p. 313). Pulmonate gastropods have tentacles used for sensing the external environment (Case in Barker 2001, p.180) (Fig 1). Terrestrial gastropod mollusks have no acoustic sense and limited visual perception (Case in Barker 2001, p.180). Olfaction takes place at sensitive olfactory organs located at the tips of each of the four tentacles and is the principal sense of perception at a distance (Case in Barker 2001, p.180).

1 shell

stomach tentacle

salivary duct

foot

genital pore dart sac

Figure 1: Basic anatomy of a Pulmonate snail. Illustration: Creative Commons Attribution-Share Alike 4.0 International, 3.0 Unported, 2.5 Generic, 2.0 Generic, and 1.0 Generic license.

Order/Clade Stylommatophora, Family Polygridae Stylommatophorid Pulmonates include most terrestrial species, have two pairs of retractable tentacles, and possess calcareous shells, except in some slugs where the shell is reduced or absent (Ruppert et al 2004, p. 342 - 343). The shells of Stylommatophorid species are not as calcified as is typical of gastropods found in the marine environment, but calcium viability remains a limiting resources for snails and often influence there distribution on the landscape Ruppert et al 2004, p. 343). The family Polygridae is composed completely of snails. Typically these snails are small to large in size with well reflected lip margins and often apertural teeth (Burke 2013, p.150). This family lacks the very interesting reproductive stimulating feature “the love dart” and contains a singular muscular band to retract eyes and pharynx (Pilsbry, 1940, p 575) Genus Cryptomastix The genus Cryptomastix was elevated from a subgenus of the large Triodpsis genus based on reproductive anatomy and distribution (Pilsbry 1939, pp. vii, xvii). Cryptomastix snails cannot be differentiated from others in the Polygyrid family solely on the basis of morphological shell characters. Instead, the details of the male anatomy must be examined. Species of the genus Cryptomastix are restricted the Pacific Northwest (Burke 2013, p. 156). Cryptomastix are described as being medium to moderately large snails with sub globose to low- conic shells and (as mature individuals) a distantly flared or reflected apertural lip and presence of apertural denticle(s) (0 – 3) (Burke 2013 p. 156, Forsyth 2004, 151 - 154). Juveniles may have some periostracal hairs, but are typically absent in mature specimens with the exception of Cryptomastix germana (Burke 2013, p 150, 158). Currently the genus is represented by 8 species: C. devia, C. germana (2 subspecies), C. vancouverinsulae, C. hardfordiana, C. hendersoni, C. magnidentata, C. mullani (7 subspecies), C. populi, and C. sanburni (Burke 2013, p 37-38, 156 -169). Note that C. hendersoni, the Columbia Oregonian was also identified in the 2008 petition to list 32 mollusk species and had a positive 90-day finding. This species has been identified for review in 2022 (Curry et al, 2008 p. 9, 76 FR 61828, USFWS 2019, p. 13)

2 1

2 Appendix II: Maxent model for Puget Oregonian 3 4 This appendix contains some analysis of the Maxent model for CRDE, created Thu May 30 11:37:38 PDT 5 2019 using Maxent version 3.4.1. If you would like to do further analyses, the raw data used here is 6 linked to at the end of this page. 7 8 9 10 Analysis of omission/commission

11 The following picture shows the omission rate and predicted area as a function of the cumulative threshold. 12 The omission rate is calculated both on the training presence records, and (if test data are used) on the test 13 records. The omission rate should be close to the predicted omission, because of the definition of the 14 cumulative threshold. 15 Omission and Predicted Area for CRDE I Fraction of background predicted • 1.0 Omission on training samples • r1 Omission on test samples • 0.9 / Predicted omission • 0.8 '.J .. 0.7 ~ :::, g; 0 6 /? r .; /;V r J C: i 0.5 / ~ /r r LL 0.4 / J/ 0.3 \ 0.2 I/ r , .-.::; ~ I IA ...... 0.1 ., # 0.0

0 10 20 30 40 50 60 70 80 90 100 Cumulative threshold 16 17 The next picture is the receiver operating characteristic (ROC) curve for the same data. Note that the 18 specificity is defined using predicted area, rather than true commission (see the paper by Phillips, Anderson 19 and Schapire cited on the help page for discussion of what this means). This implies that the maximum 20 achievable AUC is less than 1. If test data is drawn from the Maxent distribution itself, then the maximum 21 possible test AUC would be 0.920 rather than 1; in practice the test AUC may exceed this bound.

Sensitivity vs. 1 - Specificity for CRDE

I Training data (AUC = 0.934) ■ 1.0 ~ Test data (AUC = 0.936) ■ J / Random Prediction (AUC = 0.5) ■ 0.9 I ll / 0.8 - I/ m 1ii o:: 0.7 J / C 0 ~ 0.6 I/ E 0 , 0.5 / .;:;- / ~ 0.4 Cf) C ~ 0.3 /

0.2 /

0.1 /

0.0 V

0.0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1. 0 1 - Specifi city (Fracti onal Predicted Area)

Some common thresholds and corresponding omission rates are as follows. If test data are available, binomial probabilities are calculated exactly if the number of test samples is at most 25, otherwise using a normal approximation to the binomial. These are 1-sided p-values for the null hypothesis that test points are predicted no better than by a random prediction with the same fractional predicted area. The "Balance" threshold minimizes 6 * training omission rate + .04 * cumulative threshold + 1.6 * fractional predicted area.

Fractional Training Test Cumulative Logistic Description predicted omission omission P-value threshold threshold area rate rate 1.017E- 1.000 0.021 Fixed cumulative value 1 0.398 0.000 0.000 8 5.247E- 5.000 0.110 Fixed cumulative value 5 0.249 0.027 0.050 11 4.079E- 10.000 0.212 Fixed cumulative value 10 0.192 0.074 0.050 13 1.432E- 1.464 0.031 Minimum training presence 0.361 0.000 0.000 9 1.168E- 11.948 0.234 10 percentile training presence 0.178 0.096 0.150 10 Equal training sensitivity and 1.056E- 15.792 0.268 0.154 0.154 0.150 specificity 11 8.955 0.196 Maximum training sensitivity 0.201 0.053 0.050 9.234E- plus specificity 13

Equal test sensitivity and 7.047E- 16.437 0.271 0.150 0.160 0.150 specificity 12 Maximum test sensitivity plus 2.933E- 10.422 0.217 0.189 0.080 0.050 specificity 13 Balance training omission, 3.285E- 2.832 0.061 predicted area and threshold 0.299 0.005 0.000 11 value Equate entropy of thresholded 2.718E- 7.734 0.168 0.213 0.043 0.050 and original distributions 12

Pictures of the model

This is a representation of the Maxent model for CRDE. Warmer colors show areas with better predicted conditions. White dots show the presence locations used for training, while violet dots show test locations. Click on the image for a full-size version.

(A link to the Explain tool was not made for this model. The model uses product features, while the Explain tool can only be used for additive models.)

Response curves

These curves show how each environmental variable affects the Maxent prediction. The curves show how the predicted probability of presence changes as each environmental variable is varied, keeping all other environmental variables at their average sample value. Click on a response curve to see a larger version. Note that the curves can be hard to interpret if you have strongly correlated variables, as the model may depend on the correlations in ways that are not evident in the curves. In other words, the curves show the marginal effect of changing exactly one variable, whereas the model may take advantage of sets of variables changing together.

elevation ncM precip 1.0 1.0 1.0 - ~ 0.5 _/ 0.0 ., 2602 -2000 9886 0 5243. 15

slope tpic tpif 1.0 1.0 i.o~=i======i==i

0.5 0.5 0.5 Ld:::::::====D --...... - 0.0 0.0 0.0 ~ =t======l=:I 0 55.466 -481.31 447.543 -47.766 71 .04

In contrast to the above marginal response curves, each of the following curves represents a different model, namely, a Maxent model created using only the corresponding variable. These plots reflect the dependence of predicted suitability both on the selected variable and on dependencies induced by correlations between the selected variable and other variables. They may be easier to interpret if there are strong correlations between variables.

elevation precip 1.0 1.0

0.5 , - - \ 0.0 ., 2602 -2000 9886 0 5243. 15

slope 1.0

0.5 - V 0.0 0 55.466 -481.31 447.543 -47.766 71 .04

Analysis of variable contributions

The following table gives estimates of relative contributions of the environmental variables to the Maxent model. To determine the first estimate, in each iteration of the training algorithm, the increase in regularized gain is added to the contribution of the corresponding variable, or subtracted from it if the change to the absolute value of lambda is negative. For the second estimate, for each environmental variable in turn, the values of that variable on training presence and background data are randomly permuted. The model is reevaluated on the permuted data, and the resulting drop in training AUC is shown in the table, normalized to percentages. As with the variable jackknife, variable contributions should be interpreted with caution when the predictor variables are correlated.

Variable Percent contribution Permutation importance ndvi 48.1 12.4 elevation 24.4 45.7 precip 15 32.2 tpic 10.7 9.1 slope 1.5 0.4 tpif 0.3 0.2

The following picture shows the results of the jackknife test of variable importance. The environmental variable with highest gain when used in isolation is precip, which therefore appears to have the most useful information by itself. The environmental variable that decreases the gain the most when it is omitted is elevation, which therefore appears to have the most information that isn't present in the other variables.

Jackknife of regularized training gain for CRDE

elevati on W ith out vari ab le ■ Q) W ith only vari ab le ■ ..l:l ·c::tll ndvi W ith all vari abl es ■ tll > prec ip ctll Q) E slope C 2 ·;:: tpic C w

0.0 0.2 0.4 0.6 0.8 1.0 1. 2 1.4 1.6 reg ul arized training gain

The next picture shows the same jackknife test, using test gain instead of training gain. Note that conclusions about which variables are most important can change, now that we're looking at test data.

Jackknife of test gain for CRDE

elevati on - W ith out vari ab le ■ ~ W ith only vari ab le ■ ..l:l - tll ndvi ·c:: W ith all vari abl es ■ rn > prec ip crn Q) - E slope C 2 - ·;:: tpic C w tpif -

0.0 0.2 0.4 0.6 0.8 1.0 1. 2 1.4 1. 6 1.8 test gain

Lastly, we have the same jackknife test, using AUC on test data.

Jackknife of AUC for CRDE

elevation With out vari ab le ■ Q) ■ .s::, With only vari ab le rn ndvi With all vari ables ■ ~ rn > pre cip crn Q) E slope C 2 ·;;: tpi c C w tpif

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 AUC

Raw data outputs and control parameters

The data used in the above analysis is contained in the next links. Please see the Help button for more information on these. The model applied to the training environmental layers The coefficients of the model The omission and predicted area for varying cumulative and raw thresholds The prediction strength at the training and (optionally) test presence sites Results for all species modeled in the same Maxent run, with summary statistics and (optionally) jackknife results

Regularized training gain is 1.548, training AUC is 0.934, unregularized training gain is 1.724. Unregularized test gain is 1.770. Test AUC is 0.936, standard deviation is 0.016 (calculated as in DeLong, DeLong & Clarke-Pearson 1988, equation 2). Algorithm terminated after 500 iterations (0 seconds).

The follow settings were used during the run: 188 presence records used for training, 20 for testing. 2688 points used to determine the Maxent distribution (background points and presence points). Environmental layers used (all continuous): elevation ndvi precip slope tpic tpif Regularization values: linear/quadratic/product: 0.050, categorical: 0.250, threshold: 1.000, hinge: 0.500 Feature types used: product linear quadratic responsecurves: true jackknife: true outputformat: logistic outputdirectory: C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\output2 samplesfile: C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\pts\CRDE_loc.csv environmentallayers: C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\vars randomtestpoints: 10 maximumbackground: 2500 hinge: false autofeature: false threads: 4

Command line used:

Command line to repeat this species model: java density.MaxEnt nowarnings noprefixes -E "" - E CRDE responsecurves jackknife outputformat=logistic outputdirectory=C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\output2 samplesfile=C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\pts\CRDE_loc.csv environmentallayers=C:\Users\ale\Documents\GIStemp\Puget_OR\maxent\vars randomtestpoints=10 maximumbackground=2500 nohinge noautofeature threads=4 -N curvature

Appendix III – Resiliency Unit Current and Future Condition Scoring

To assess the condition of each resiliency unit, we assessed the state of occurrence data (temporal relevance and number) and habitat in and around each unit. We assigned scores to each of these the four factors for each resiliency unit, as shown in the table below.

Score Temporal Relevance Number of Occurrences Habitat Quality Habitat Connectivity 3 Contemporary Common High Suitability High Connectivity 2 Late Century Uncommon Suitable Moderate Connectivity 1 Mid-Century Rare Low Suitability Low Connectivity 0 Historic Not Viable/Unknown Not Suitable Isolated

To arrive at an overall condition for each resiliency unit, we averaged the four scores for each resiliency unit to obtain an overall score. The overall condition categories and their corresponding score thresholds are shown in the table below.

Bound Unknown Low Moderate High upper 0.75 1.50 2.25 3.00 lower 0.00 0.76 1.51 2.26

For scenarios in which we expect declines in habitat quality and/or connectivity (Scenarios 2, 3, and 4), we analyzed the entire range of possible declines. For each these scenarios, we calculated scores for both mild declines and severe declines. The condition scoring tables for the present each of the future scenarios are shown below.

Current Condition and Scenario 1 (No Change) Scores for all Resiliency Units

% Resiliency % Habitat count per No. of Habitat around HUCCODE HUC_NAME Sub-basin Resiliency Unit Name observations Year in unit unit 17090011 Clackamas 1 Clackamas-1 2 1999 73% 60% 17090011 Clackamas 2 Clackamas-2 1 1999 96% 93% 17110016 Deschutes 3 Deschutes-3 1 1937 94% 98% 17110013 Duwamish 4 Duwamish-4 1 1980 80% 68% 17110013 Duwamish 1 Duwamish-1 1 2007 60% 52% 17110018 Hood Canal 1 Hood Canal-1 1 1999 30% 24% 17110012 Lake Washington 1 Lake Washington-1 1 1994 64% 77% 17110012 Lake Washington 2 Lake Washington-2 1 0 71% 51% 17110012 Lake Washington 3 Lake Washington-3 2 1952 48% 54% 17080002 Lewis 1 Lewis-1 4 1941 100% 95% 17080002 Lewis 2 Lewis-2 1 0 100% 100% 17080002 Lewis 3 Lewis-3 1 2001 4% 5% 17080002 Lewis 4 Lewis-4 1 1998 91% 62% 17080002 Lewis 5 Lewis-5 1 1996 46% 21% 17080002 Lewis 6 Lewis-6 1 2002 78% 73% Lower Columbia- Lower Columbia- 17080003 Clatskanie 1 Clatskanie-1 1 2004 98% 91% Lower Columbia- Lower Columbia- 17080003 Clatskanie 2 Clatskanie-2 1 0 92% 97% Lower Columbia- Lower Columbia- 17080001 Sandy 1 Sandy-1 1 0 36% 28% 17080005 Lower Cowlitz 1 Lower Cowlitz-1 1 0 98% 94% 17080005 Lower Cowlitz 2 Lower Cowlitz-2 2 2006 96% 86% 17080005 Lower Cowlitz 3 Lower Cowlitz-3 1 2008 98% 95% 17090012 Lower Willamette 1 Lower Willamette-1 1 0 93% 90% 17090012 Lower Willamette 2 Lower Willamette-2 2 1944 100% 97% 17090012 Lower Willamette 3 Lower Willamette-3 2 1946 93% 94% Middle Columbia- Middle Columbia- 17070105 Hood 1 Hood-1 1 0 0% 0% Middle Columbia- Middle Columbia- 17070105 Hood 2 Hood-2 2 1999 92% 76% Middle Columbia- Middle Columbia- 17070105 Hood 3 Hood-3 2 1906 99% 93% Middle Columbia- Middle Columbia- 17070105 Hood 4 Hood-4 4 2006 100% 97% Middle Columbia- Middle Columbia- 17070105 Hood 5 Hood-5 2 1999 100% 86% Middle Columbia- Middle Columbia- 17070105 Hood 6 Hood-6 2 1998 85% 65% Middle Columbia- Middle Columbia- 17070105 Hood 7 Hood-7 3 2001 85% 65% 17030002 Naches 1 Naches-1 1 0 0% 0% 17110015 Nisqually 1 Nisqually-1 1 2018 100% 99%

17110015 Nisqually 2 Nisqually-2 1 1990 98% 99% 17110015 Nisqually 3 Nisqually-3 2 1983 90% 87% 17110015 Nisqually 4 Nisqually-4 1 1952 68% 79% 17110019 Puget Sound 1 Puget Sound-1 3 2004 82% 89% 17110019 Puget Sound 2 Puget Sound-2 2 1892 10% 31% 17110019 Puget Sound 3 Puget Sound-3 1 0 99% 89% 17110019 Puget Sound 4 Puget Sound-4 1 0 42% 39% 17110019 Puget Sound 5 Puget Sound-5 2 1929 1% 9% 17110019 Puget Sound 6 Puget Sound-6 4 2011 5% 18% 17110019 Puget Sound 7 Puget Sound-7 1 2004 95% 91% 17110014 Puyallup 1 Puyallup-1 1 1983 94% 96% 17110014 Puyallup 2 Puyallup-2 1 1995 100% 100% 17110014 Puyallup 3 Puyallup-3 1 2006 100% 100% 17110006 Sauk 1 Sauk-1 1 2004 100% 84% 17110010 Snoqualmie 1 Snoqualmie-1 6 2005 82% 72% 17110010 Snoqualmie 2 Snoqualmie-2 4 1966 68% 72% 17110008 Stillaguamish 1 Stillaguamish-1 1 2004 48% 23% 17090010 Tualatin 1 Tualatin-1 2 1910 97% 99% 17100103 Upper Chehalis 1 Upper Chehalis-1 1 1997 72% 64% 17080004 Upper Cowlitz 1 Upper Cowlitz-1 1 1998 70% 41% 17080004 Upper Cowlitz 2 Upper Cowlitz-2 20 2010 97% 72% 17080004 Upper Cowlitz 3 Upper Cowlitz-3 7 2006 94% 90% 17080004 Upper Cowlitz 4 Upper Cowlitz-4 2 2003 83% 80% 17080004 Upper Cowlitz 5 Upper Cowlitz-5 3 2000 33% 49% 17080004 Upper Cowlitz 6 Upper Cowlitz-6 46 2010 90% 78% 17080004 Upper Cowlitz 7 Upper Cowlitz-7 36 2009 94% 81% 17080004 Upper Cowlitz 8 Upper Cowlitz-8 24 2007 98% 91% 17080004 Upper Cowlitz 9 Upper Cowlitz-9 1 2002 84% 82% 17080004 Upper Cowlitz 10 Upper Cowlitz-10 1 2001 99% 100% 17080004 Upper Cowlitz 11 Upper Cowlitz-11 6 2001 100% 99% 17080004 Upper Cowlitz 12 Upper Cowlitz-12 2 2000 90% 90% 17080004 Upper Cowlitz 13 Upper Cowlitz-13 7 2001 97% 82% 17080004 Upper Cowlitz 14 Upper Cowlitz-14 1 1998 86% 80% 17080004 Upper Cowlitz 15 Upper Cowlitz-15 3 2001 91% 79% 17080004 Upper Cowlitz 16 Upper Cowlitz-16 4 1999 93% 79% 17080004 Upper Cowlitz 17 Upper Cowlitz-17 1 2009 97% 89% 17030001 Upper Yakima 1 Upper Yakima-1 1 2000 0% 19% 17090008 Yamhill 1 Yamhill-1 2 2001 77% 90% 17090008 Yamhill 2 Yamhill-2 1 2001 83% 90% 17090008 Yamhill 3 Yamhill-3 1 2013 100% 98% 17090008 Yamhill 4 Yamhill-4 1 2017 86% 96%

Current Condition and Scenario 1 (No Change) Scores for Resiliency Units evaluated

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 3 3 2.50 high Middle Columbia-Hood-7 3 1 3 2 2.25 moderate Puget Sound-1 3 1 3 3 2.50 high Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 3 2 2.25 moderate Snoqualmie-2 1 1 2 2 1.50 low Upper Cowlitz-11 3 1 3 3 2.50 high Upper Cowlitz-13 3 1 3 3 2.50 high Upper Cowlitz-15 3 1 3 3 2.50 high Upper Cowlitz-16 3 1 3 3 2.50 high Upper Cowlitz-2 3 3 3 2 2.75 high Upper Cowlitz-3 3 3 3 3 3.00 high Upper Cowlitz-5 3 1 1 1 1.50 low Upper Cowlitz-6 3 3 3 3 3.00 high Upper Cowlitz-7 3 3 3 3 3.00 high Scenario 2 (Decline in Habitat Connectivity) Scores

Mild Decline - Habitat Connectivity scores decrease by 1

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 3 2 2.25 moderate Middle Columbia-Hood-7 3 1 3 1 2.00 moderate Puget Sound-1 3 1 3 2 2.25 moderate Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 3 1 2.00 moderate Snoqualmie-2 1 1 2 1 1.25 low Upper Cowlitz-11 3 1 3 2 2.25 moderate Upper Cowlitz-13 3 1 3 2 2.25 moderate Upper Cowlitz-15 3 1 3 2 2.25 moderate Upper Cowlitz-16 3 1 3 2 2.25 moderate Upper Cowlitz-2 3 3 3 1 2.50 high Upper Cowlitz-3 3 3 3 2 2.75 high Upper Cowlitz-5 3 1 1 0 1.25 low Upper Cowlitz-6 3 3 3 2 2.75 high Upper Cowlitz-7 3 3 3 2 2.75 high

Severe Decline - Habitat Connectivity scores decrease to 0

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 3 0 1.75 moderate Middle Columbia-Hood-7 3 1 3 0 1.75 moderate Puget Sound-1 3 1 3 0 1.75 moderate Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 3 0 1.75 moderate Snoqualmie-2 1 1 2 0 1.00 low Upper Cowlitz-11 3 1 3 0 1.75 moderate Upper Cowlitz-13 3 1 3 0 1.75 moderate Upper Cowlitz-15 3 1 3 0 1.75 moderate Upper Cowlitz-16 3 1 3 0 1.75 moderate Upper Cowlitz-2 3 3 3 0 2.25 moderate Upper Cowlitz-3 3 3 3 0 2.25 moderate Upper Cowlitz-5 3 1 1 0 1.25 low Upper Cowlitz-6 3 3 3 0 2.25 moderate Upper Cowlitz-7 3 3 3 0 2.25 moderate

Scenario 3 (Decline in Habitat Quality) Scores

Mild Decline - Habitat Quality scores decrease by 1

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 2 3 2.25 moderate Middle Columbia-Hood-7 3 1 2 2 2.00 moderate Puget Sound-1 3 1 2 3 2.25 moderate Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 2 2 2.00 moderate Snoqualmie-2 1 1 1 2 1.25 low Upper Cowlitz-11 3 1 2 3 2.25 moderate Upper Cowlitz-13 3 1 2 3 2.25 moderate Upper Cowlitz-15 3 1 2 3 2.25 moderate Upper Cowlitz-16 3 1 2 3 2.25 moderate Upper Cowlitz-2 3 3 2 2 2.50 high Upper Cowlitz-3 3 3 2 3 2.75 high Upper Cowlitz-5 3 1 0 1 1.25 low Upper Cowlitz-6 3 3 2 3 2.75 high Upper Cowlitz-7 3 3 2 3 2.75 high

Severe Decline - Habitat Quality scores decrease to 0

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 0 3 1.75 moderate Middle Columbia-Hood-7 3 1 0 2 1.50 low Puget Sound-1 3 1 0 3 1.75 moderate Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 0 2 1.50 low Snoqualmie-2 1 1 0 2 1.00 low Upper Cowlitz-11 3 1 0 3 1.75 moderate Upper Cowlitz-13 3 1 0 3 1.75 moderate Upper Cowlitz-15 3 1 0 3 1.75 moderate Upper Cowlitz-16 3 1 0 3 1.75 moderate Upper Cowlitz-2 3 3 0 2 2.00 moderate Upper Cowlitz-3 3 3 0 3 2.25 moderate Upper Cowlitz-5 3 1 0 1 1.25 low Upper Cowlitz-6 3 3 0 3 2.25 moderate Upper Cowlitz-7 3 3 0 3 2.25 moderate

Scenario 4 (Decline in Habitat Quality and Connectivity) Scores

Mild Decline - Habitat Quality and Connectivity scores decrease by 1

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 2 2 2.00 moderate Middle Columbia-Hood-7 3 1 2 1 1.75 moderate Puget Sound-1 3 1 2 2 2.00 moderate Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 2 1 1.75 moderate Snoqualmie-2 1 1 1 1 1.00 low Upper Cowlitz-11 3 1 2 2 2.00 moderate Upper Cowlitz-13 3 1 2 2 2.00 moderate Upper Cowlitz-15 3 1 2 2 2.00 moderate Upper Cowlitz-16 3 1 2 2 2.00 moderate Upper Cowlitz-2 3 3 2 1 2.25 moderate Upper Cowlitz-3 3 3 2 2 2.50 high Upper Cowlitz-5 3 1 0 0 1.00 low Upper Cowlitz-6 3 3 2 2 2.50 high Upper Cowlitz-7 3 3 2 2 2.50 high

Severe Decline - Habitat Quality and Connectivity scores decrease to 0

Resiliency Unit Name Temporal Number of Habitat Habitat Average Condition Relevance Occurrences Quality Connectivity Score Middle Columbia-Hood-4 3 1 0 0 1.00 low Middle Columbia-Hood-7 3 1 0 0 1.00 low Puget Sound-1 3 1 0 0 1.00 low Puget Sound-6 3 1 0 0 1.00 low Snoqualmie-1 3 1 0 0 1.00 low Snoqualmie-2 1 1 0 0 0.50 low Upper Cowlitz-11 3 1 0 0 1.00 low Upper Cowlitz-13 3 1 0 0 1.00 low Upper Cowlitz-15 3 1 0 0 1.00 low Upper Cowlitz-16 3 1 0 0 1.00 low Upper Cowlitz-2 3 3 0 0 1.50 low Upper Cowlitz-3 3 3 0 0 1.50 low Upper Cowlitz-5 3 1 0 0 1.00 low Upper Cowlitz-6 3 3 0 0 1.50 low Upper Cowlitz-7 3 3 0 0 1.50 low

Appendix IV Climate Change

Greenhouse gas emissions have increased at an unprecedented rate during the 20th century, resulting in global climate change; these changes have been characterized by warming atmospheric and ocean temperatures, diminishing snow and ice, and rising sea levels (Intergovernmental Panel on Climate Change (IPCC) 2014, pp. 2-3). Scientists use a variety of climate models, which include consideration of natural processes and variability, as well as various scenarios of potential levels and timing of greenhouse gas (GHG) emissions, to evaluate the causes of changes already observed and to project future changes in temperature and other climate conditions (e.g., Meehl et al. 2007, entire; Ganguly et al. 2009, pp. 11555, 15558; Prinn et al. 2011, pp. 527, 529). Combinations of models and emissions scenarios yield similar projections of increases in the most common measure of climate change, average global surface temperature (commonly known as global warming), until about 2030. Although projections of the magnitude and rate of warming differ after about 2030, the overall trajectory of all the projections is one of increased temperature through the end of this century, even for the projections based on scenarios that assume that greenhouse gas emissions will stabilize or decline. There is strong scientific support that warming will continue through the 21st century, and that the magnitude and rate of this change will be influenced substantially by the extent of greenhouse gas emissions (IPCC 2014, p. 8; Meehl et al. 2007, pp. 760–764 and 797-811; Ganguly et al. 2009, pp. 15555–15558; Prinn et al. 2011, pp. 527, 529). Climate change is likely to impact the Puget Oregonian through direct effects to individuals and populations, and indirect effects to suitable habitat. In order to incorporate these projected effects of climate change into future conditions for the Puget Oregonian, we analyzed predictions of two IPCC greenhouse gas emissions scenarios. The IPCC identifies various greenhouse gas Representative Concentration Pathways (RCPs) which take into account different scenarios of greenhouse gas emissions, atmospheric concentrations, and land use likely to unfold in the 21st century. The IPCC characterizes several potential scenarios including RCP 4.5, an intermediate emissions scenario where atmospheric CO2 concentrations are expected to equal approximately 650 ppm after the year 2100, and RCP 8.5, where emissions aggressively increase to approximately 1370 ppm CO2 after the year 2100. For comparison, current atmospheric CO2 concentrations are around 400 ppm (IPCC 2014, p. 57). For the purposes of analyzing future conditions for the Puget Oregonian, we considered one intermediate scenario that assumes moderate cuts are made to emissions (RCP 4.5), and one high emissions scenario that assumes no deviation from the current emissions trajectory (RCP 8.5). These emissions scenarios were chosen because they frame the most likely high and low boundaries of the possible future in regard to greenhouse gas emissions. We consider future scenarios using the time period from the present to the middle of the century (2040-2069) (approximately 20-50 years), which is within the timeframe most climate models make projections for, as the uncertainty of future climate response to global warming increases with time from the present (IPCC 2014, p. 59).

We analyzed the effects of climate change in areas that overlap with known Puget Oregonian populations through the middle of the century using data obtained from the Northwest Climate Toolbox, developed by members of the Applied Climate Science Lab at the University of Idaho (Northwest Climate Toolbox, 2019). In addition to past and current data, the Northwest Climate Toolbox provides modeled future projections of climate, hydrology, and fire danger, based on the effects of potential degrees of greenhouse gas emissions reported by the IPCC (IPCC, 2014, entire). Each future projection dataset we used for the purpose of analysis was a multi-model mean derived from multiple downscaled Coupled Model Intercomparison Project 5 (CMIP5) models. Though projections from individual models will vary for many reasons, the multi-model means often provide a good central estimate of the projected change (Northwest Climate Toolbox, 2019). Data and projections obtained from the Northwest Climate Toolbox provide estimates of future conditions, but may not be entirely accurate for any given site or year. Climate change within the United States has followed global trends, with the largest changes occurring in the western portion of the country, including the range of the Puget Oregonian. Between the periods of 1901-1960 and 1986-2016, average temperature in the Northwest has increased by more than 0.8°C (1.5°F). Historically (1971-2000), the annual average number of warm days (above 86°F or 30°C) have numbered 10 or below across most of the Puget Oregonian’s range, with some limited portions of the southern part of the range averaging up to 21-30 days. Under a low emissions scenario (RCP 4.5), the number of warm days is expected to increase to the range of 21-30 days by 2040-2069 in signification portions of the northern part of the snail’s range. In the southern part of the snail’s range, most locations are expected to experience between 41-50 warm days, with some locations reaching 51-60 warm days by 2040- 2069. Under a higher emissions scenario (RCP 8.5) projected increases are even more severe, with large portions of the species’ range expected to experience 41-70 warm days annually by 2040-2069 (Hegewisch et al. 2019). While annual precipitation has changed in other parts of the United States, it has remained relatively constant in the Pacific Northwest (USGCRP 2018 p. 88-89). The average annual precipitation in the range of the Puget Oregonian is expected to continue to stay constant, in both low (RCP 4.5) and high (RCP 8.5) emissions scenarios, through the 2040-2069 and 2070-2099 periods, as compared to the 1971-2000 period. However, seasonal variation in precipitation is predicted to increase over these same scenarios and time periods. Across the range of the Puget Oregonian, average winter precipitation is expected to increase by about 2.5 to 7.6 cm (1 to 3 in) under a low emissions scenario (RCP 4.5) and 5.1 to 10.2 cm (2 to 4 in) under a high emissions scenario (RCP 8.5) by 2040-2069. Meanwhile, average summer precipitation is expected to decrease, averaging up to about 5.1 cm (2 in) of decrease under a low emissions scenario (RCP 4.5) and up to about 7.6 cm (3 in) of decrease under a high emissions scenario (RCP 8.5) by 2040-2069, depending on location. Changes in expected mean precipitation during the spring and fall vary depending on location within the Puget Oregonian’s range. In the northern portion of the species’ range, spring and fall precipitation is predicted to increase by up to 5.1 cm (2 in) and about 7.6 cm (3 in) by 2040-2069, in low emissions (RCP 4.5) and high emissions (RCP 8.5) scenarios, respectively. Increases in fall precipitation are expected to affect a larger portion of the snail’s range than during the spring. In the southern portion of the species’ range, spring and

fall precipitation is predicted to decrease by up to about 5.1 cm (2 in) in both low and high emissions scenarios by 2040-2069. Decreases in spring precipitation are expected to affect a larger portion of the snail’s range than during the fall (Hegewisch et al. 2019). The predicted changes to winter precipitation within the range of the Puget Oregonian are likely to lead to other related extreme events such as floods. Floods are influenced by other factors in addition to precipitation, such as land cover, land use, and water management, and thus are more difficult to predict than precipitation. However, flooding events are generally expected to increase in frequency and intensity in the Pacific Northwest (USGCRP 2018 p. 90-91). Throughout the region, the frequency 20-year extreme precipitation events is expected to increase by about 9 percent under a low emissions scenario (RCP 4.5), and by about 11 percent under a high emissions scenario by the middle of the 21st century, relative to 1986-2015 (USGCRP 2017 p. 216-222). Flooding events will likely be further driven by the increased amount of winter precipitation falling as rain rather than snow, and the amount of rain on snow events throughout the region (USGCRP 2018 p. 90-91). Meanwhile, changes in summer precipitation and temperature within the range of the Puget Oregonian are likely to lead to drier summer conditions across the region. Annual surface soil moisture is projected to decrease across the region, as a result of decreasing summer precipitation, and higher evaporation rates due to increased temperatures. Dry summer conditions will likely be further driven by expected decreases in snowpack, and earlier snow melt in the spring (USGCRP 2018 p. 91). Historically (1971-2000), soil moisture within the eastern and western range of the snail has measured mostly between 78.7 and 101.6 cm (1.31 and 40 inches) of water contained in the upper few meters of soil, with the central portion in between averaging 40.6-76.2 cm (16-30 inches). Under a low emissions scenario (RCP 4.5), annual soil moisture is expected decrease by up to about 5.1 cm (2 inches) throughout the region by 2040- 2069, while under a high emissions scenario (RCP 8.5), annual soil moisture is expected decrease by up to about 12.7 cm (5 inches) throughout the region, over that same time period (Hegewisch et al. 2019). The conditions leading to increased summer drought, including reduced summer precipitation, increased temperatures, increased vapor pressure deficit, and earlier snowmelt, are also expected to decrease fuel moisture, and thus increase fire risk throughout the range of the Puget Oregonian. Due to decades of fire suppression efforts and a warmer drier climate, the area burned annually by wildfires across the United States has steadily increased over the past 20 years, and is expected to continue increasing (USGCRP 2018 p. 240-243). Measures such as 100-hour fuel moisture can serve as indicators for fire risk across the landscape. 100-hour fuel moisture is defined as the average amount of moisture across 100 hours in dead vegetation in the 2.5 to 7.6 cm (1- to 3-inch) diameter class available to a fire, as expressed as a percentage of the fuel’s dry weight. Historically (1971-2000), fuel moisture within the range of the snail has measured mostly between 19 and 21 percent. On average, 100-hour fuel moisture is expected to stay relatively constant or slightly increase in the region by 2040-2069, under both a low emissions scenario (RCP 4.5) and a high emissions scenario (RCP 8.5). However, the variability in 100-hour fuel moisture, including the number of days when 100-hour fuel moisture has

abnormally low values compared to historic values, is expected to increase. The number of high, very high, and extreme fire danger days (defined as days with 100-hour fuel moisture below the 20th, 10th, and 3rd percentile of historic values, respectively) are all expected to increase throughout the range of the Puget Oregonian, with higher fire risk occurring in the northern part of the range than the southern part. By 2040-2069, high, very high, and extreme fire danger days are expected to number about 79 to 90 days, 44 to 55 days, and 16 to 25 days under a low emissions scenario (RCP 4.5), respectively. Under a high emissions scenario (RCP 8.5), the number of fire danger days are expected to number 83 to 140 days, 51 to 60 days, and 16 to 30 days for those same fire danger categories by the same time period (Hegewisch et al. 2019).