Endemic Small Mammals of Southeastern Alaska: Evolutionary Diversity, Ecology,

1
Endemic Small Mammals of Southeastern Alaska
Evolutionary Diversity, Ecology, Conservation

• Winston P. Smith
• USDA Forest Service, PNW Research Station
• Forestry Sciences Laboratory
• Juneau, AK 99801-8545 USA

2
Acknowledgments
__________________________________________________
__________________ WRANGELL, THORNE BAY and CRAIG
Ranger Districts, Tongass National Forest. UAM
Joe Cook, Steve MacDonald, Chris Conroy, John
Demboski, Karen Stone, and Amy Runck. SPECIAL
THANKS Jeff Nichols, field crew leader and
analytical support Scott Gende, analytical
support and comments on presentation and
completion reports. Lillian Petershoare and
JFSL Library.
3
Background Setting

• __________________________________________________
  _______________
• Unique attributes
  - dynamic recent geological
  history - naturally
  fragmented and isolated habitat - largest NF and
  largest remaining temperate rainforest
  - spatial and temporal
  heterogeneity
• Limited information on natural history
  
• Depauperate small mammal fauna
  
• High potential for endemism
  

4
Background Planning

• __________________________________________________
  _______________
• Extensive clearcut logging since 1954
  - 50 of most
  productive forest on some islands 40
  of some watersheds
  - gt300 yrs to
  develop old forest features
• TLMP - endemic small mammals ranked as highest
  risk of extinction
• Conservation strategy
  - metapopulation framework -
  lacks empirical foundation
  

5
Land Mammal Fauna

• __________________________________________________
  _______________
• 87 islands 24 mainland localities
• 45 land mammal species - 65 small (lt10kg)
  mammal taxa 15
  endemic 11 confined 36 widely distributed, 3
  unknown
• Mammal fauna nested structure with varying
  genetic divergence
  - colonization rather than
  extinction - significant relationship
  between isolation and species richness
  - competition influenced similar species
  (e.g., bears, shrews)

6
Land Mammal Fauna

• __________________________________________________
  _______________
• - neo-endemics (flying squirrel) and paleo-
  endemics (marten) - number of endemics on
  outer islands suggests glacial refugia -
  some taxa have affinities with eastern NA
  forms (e.g., flying squirrel, marten)
• Most species Upper Lynn Canal
• Most endemics Mainland Subregion
• Ermine show highest degree of endemism with 5
  subspecies representing 3 clades. (map)

7
Mustela erminea Distribution in SE
Alaska MacDonald and Cook 1996
M. e. arctica
Juneau

M. e. salva
CANADA
N
M. e. alascensis
M. e. celenda
M. e. seclusa
.
8
Peromyscus keeni1 Distribution in SE
Alaska MacDonald and Cook 1996
Island Endemics
P. k. algidus
P. k. sitkensis
P. k. macrorhinus

Juneau
CANADA
N
P. k. hylaeus
.
P. k. oceanicus
1Hogan et al. 1993
9
Island Biogeography
Processes colonization extinction
Variables island size, distance from mainland,
vagility
island
island
immigration
island
island
Continent
immigration
island
island
island
island
10
Community Dynamics and Structure
S 4
S 7
island
island
Continent
S 9
S 3
island
Species Richness (S)
island
S 12
S 14
equilibrium
Colonization
island
island
Extinction
S 5
S 3
island
Island area
island
11
Faunal Extinctions
island
island
N
Refugia
island

island
Glacial Advance
Continent
island
island
island
island
12
Regional Colonization
island
island
N
Refugia
Glacial Retreat
island

island
Continent
island
island
island
island
13
Southern red-backed vole
Photo by P. Myers
14
Natural History Red-backed vole

• _______________________________________________
  ___________
• Nearctic Distribution - in SE Alaska, southern
  mainland and nearshore islands
• Mesic forest habitat specialist
• Omnivorous, but primarily eats fungi
  (mycophagous) in the Pacific Northwest
• Sensitive to overstory removal and fire in
  western coniferous forests
• Influenced by landscape context, but little
  evidence of edge effects.
• (map)

15
Clethrionomys gapperi Distribution in SE
Alaska MacDonald and Cook 1996
Island Endemics
Juneau

CANADA
N
C. g. stikinensis
C. g. wrangeli
C. g. solus
C. g. phaeus
C. g. saturatus
16
Prince of Wales Flying Squirrel

• Markedly different from PNW populations
• Nocturnal, active year-
• round
• 1 litter (2-4 young)/year
• Reputed old-growth
• habitat specialist
• Mycophagist in PNW
• Vulnerable to isolation
• in managed landscapes
• (map)

17
Glaucomys sabrinus Distribution in SE
Alaska MacDonald and Cook 1996
G. s. zaphaeus
G. s. alpinus
Juneau


CANADA
N
G. s. griseifrons
.
18
Research Objectives

• ___________________________________
• Estimate density of red-backed voles and flying
  squirrels among habitats
• Contrast seasonal abundance, age and sex
  composition, body condition, survival, and
  reproductive condition of voles and squirrels
  among habitats
• Examine habitat use and density relative to
  microsite and stand-level vegetative and
  structural features.

19
Study Area (voles)

• ______________________________________________
  __
• Wrangell Island
• 4 habitats
• - 3 unmanaged habitats in largely unmanaged
  landscapes
• - thinned young growth stands within
• watershed with 40 POG clearcut .

20
Gap-Phase Old-Growth Forest
Structurally heterogeneous, vertically and
horizontally - large (gt150 cm), old (gt800 yr)
trees.
Northerly exposure
lt500 m elevation
Fine scale disturbance
Sitka spruce/ western hemlock
21
Multi-Cohort Old-Growth Forest
Southerly exposure Catastrophic disturbance
per 100-200 yr Heterogeneous stands of
even-aged patches Uniform diameter, dense
canopy forest 100-102 ha.
22
Unmanaged Peatland-scrub/ Mixed-conifer Forest
Structurally complex, biologically diverse
communities.
Spatially heterogeneous from open muskeg to
forest over a scale of 101 102 m.
2 replicates in 1 watershed.
23
Thinned Young Growth (25 yr-old) -pre-commercial
thinning 2-3 yr prior to study
initiation -2 replicates within the same
watershed.
Dense understory of herbaceous and woody plants
and slash.
24
Study Design

• ______________________________________
• Two replicates each of 4 habitats
• Replicates within the same watershed but gt800 m
  apart
• 1-ha grid (11 X 11) 8 assessment lines per
  replicate with 2 traps per station (n 242)
• Live trap spring 1999-2000 and early autumn
  1998-2000.

25
Study Area (squirrels)

• ______________________________________________
  ___
• North-Central Prince of Wales Island
• 2 habitats 1) POG (mostly gap-phase) and 2)
  peatland-scrub/mixed-conifer
• 1) two ends of a continuum of forest cover
• 2) peatland-scrub/mixed-conifer - little
  commercial, but potential ecological value.

26
Study Design

• ____________________________________________
• Three replicates each of 2 habitats in largely
  unmanaged landscapes
• Replicates in the same watershed but gt1 km apart
  
• 13-ha grid (10 X 10 array of traps)
• Live trap spring and early autumn 1998-2000.

27
RESULTS Demography
28
Mean Effective Area Sampled Wrangell
Red-backed Voles

• First estimates for red-backed voles, which
  precludes comparison of density with earlier
  studies
• Effective area sample was an order of magnitude
  larger than grid
• Effective area sampled was significantly greater
  in gap-phase old-growth than in other habitats
  relative comparisons among habitats are therefore
  invalid.

29
Wrangell Red-backed Vole Density Spring 1999
- 2000

• Density higher in 1999 than 2000 with significant
  differences in multi-cohort and gap-phase
  old-growth
• Density lower in peatland/mixed-conifer in both
  years but significant only in 1999.

30
Wrangell Red-backed Vole Density Autumn 1998
- 2000

• Density higher in 1998 than 1999 or 2000 with
  significant differences in all habitats
• Peatland/mixed-conifer consistently had lowest
  densities of voles with significant differences
  in 1999 and 2000
• Some evidence that thinned YG may serve as a
  habitat sink.

31
Wrangell Red-backed Vole Population Attributes
among Habitats

• Age and sex ratios were similar among habitats
• Summer survival and percentage of reproductive
  females were significantly lower in YG than other
  habitats winter survival also was lower but not
  statistically significant.

32
Prince of Wales Flying Squirrel Seasonal
Movements

• Mean maximum distance moved was similar between
  habitats and seasons, averaging about 110 yards
• Home ranges also were similar between seasons and
  and habitats and ranged from about 5.5 acres to
  about 9.5 acres.

33
POW Flying Squirrel Density Spring 1999 - 2000
34
POW Flying Squirrel Density Autumn 1998 - 2000
Reproductive
35
POW Flying Squirrel Survival and Productivity

36
Densities in the Pacific Northwest
37
Why?

• Hypotheses
• Competitive release
• Reduced predation pressure
• Diet

38
RESULTS Habitat Relations
39
Logistic Regression Model Spring
1999-2000Red-backed Voles

• Deciduous shrub cover () within 1.5 meters of
  the forest floor was the most significant
  variable influencing vole microhabitat selection
  overall - by a factor of 2 in multi-cohort
  OG and 3 in peatland/mixed-conifer.
• Moss cover (-), density of stumps () and density
  of small snags and saplings () were correlates
  of microhabitat use in gap-phase, thinned YG, and
  peatland/mixed-conifer.

40
Logistic Regression Model - Autumn
1999-2000Red-backed Voles

• Deciduous cover again had the greatest influence
  on microhabitat selection overall - cover
  between 0.3 1.5 m had 2-fold (-) in
  gap-phase - cover lt 0.3 m had 5-fold () in
  gap- phase and 5-fold (-) in peatland/mixed-
  conifer.

41
Discriminant Model- Spring 1999-2000Red-backed
Vole
Model Multivariate Factor
Gap-Phase N/A

Multi-cohort N/A

Thinned YG N/A

Peatland-MC Moss, sapling, deciduous shrub
42
Discriminant Model- Autumn 1999-2000Red-backed
Vole
Model Multivariate Factor
Gap-Phase Vaccinium deciduous shrub

Multi-cohort N/A

Thinned YG Vaccinium moss

Peatland-MC Moss, sapling, deciduous shrub
43
Red-backed Vole Density Decayed Downed Wood
Explained about 90 of variation in density
Vole density (ha)
Decay IV (volume/ha)
44
Red-backed Vole Density Conifer Seedling Cover
Explained about 85 of variation in vole density
Vole density (ha)
Conifer cover () lt30 cm
45
Habitat Correlates of Density Red-backed Vole

• Spring 1999-2000
• ________________________________
• Decayed wood class IV ()
• Soft snags 10-49 cm dbh (-)
• Coarse woody debris ()
• Decayed wood class III ()
• Conifer cover 0.3 1.5 m (-)
• Conifer cover lt0.3 m (-)
• Water ground cover ()
• Moss ground cover (-)
• Trees 10-49 cm dbh (-)

• Autumn 1999-2000
• ________________________________
• Decayed wood class IV ()
• Soft snags 10-49 cm dbh (-)
• Coarse woody debris ()
• Decayed wood class III ()
• Conifer cover 0.3 1.5 m (-)
• Conifer cover lt0.3 m (-)
• Trees 5-10 cm dbh (-)

46
Logistic Regression ModelNorthern Flying
Squirrel

• Density of trees gt74 cm dbh and cover of
  Vaccinium most influenced microhabitat use during
  spring and autumn - in peatland/mixed-conifer
  large tree density increased capture probability
  by a factor of 3 during spring and 17 in autumn
• Most influential habitat feature in gap-phase OG
  was ground cover of water was inversely
  correlated with microhabitat use.

47
Discriminant Function ModelNorthern Flying
Squirrel
Model Multivariate Factor
Spring
Peatland-MC Understory cover/structure
Large tree/ snag density

Upland-OG N/A

Autumn
Peatland-MC Large tree/ snag density

Upland-OG Vaccinium
48
Glaucomys sabrinus Density and Live Trees
gt74 cm DBH
Explained about 65 of variation in squirrel
density
Squirrels/ha
Trees gt74 cm dbh/ha
49
Ecological Correlates of Density Northern
Flying Squirrel

• Spring 1998-2000
• _____________________________________________
• Moss ground cover ()
• Decayed wood class I ()
• Decayed wood class I ()
• Decayed wood class IV ()

• Autumn 1998-2000
• __________________________________________________
  ____
• Trees gt74 cm dbh ()
• Trees 5-10 cm dbh (-)
• Trees 10-49 cm dbh (-)

50
CONCLUSIONS
51
Endemism

• ____________________________________
• Current taxonomy under-represents diversity
  complexity of mammal fauna
• High likelihood of undocumented endemic taxa,
  especially on mainland
• Outer islands show greater divergence and may
  contribute more genetic diversity
• POW flying squirrel differs markedly from
  northern flying squirrel in PNW
• Habitat loss will increase extinction risks,
  especially endemics of small islands.

52
Population Ecology

• ____________________________________
• Peatland-scrub/mixed conifer may support breeding
  populations of flying squirrels, but not
  red-backed voles
• Small HCAs likely large enough to support
  breeding populations of flying squirrels
• Voles occur in thinned YG, but it may function
  as a habitat sink
• Vole captures most often correlated with
  deciduous cover, but relationship and ecological
  impact varied among habitats -
• CANNOT ignore habitat context

53
Population Ecology

• ____________________________________
• Vaccinium - major component of factors
  correlated with autumn vole captures in Young
  Growth and Gap-Phase habitats
• Moss in YG during autumn may reflect vole needs
  for moist microsite conditions
• POG is primary habitat of POW flying squirrel,
  which may have a more general lifestyle than
  populations in PNW
• Stand level - vole density directly related to
  coarse woody debris and decayed downed wood,
  inversely related to conifer cover and dead
  saplings in the understory

54
Population Ecology

• ____________________________________
• Flying squirrel captures most influenced by large
  tree and snag density and Vaccinium cover in
  peatland/mixed-conifer where it may be limiting
• Stand level - squirrel density related to large
  tree density and decayed wood
• Inferences limited because of annual population
  variability, its influence on habitat use, and
  limited duration of study.

55
Management Implications

• ____________________________________
• Select-harvest of POG likely have minimal impacts
  to flying squirrel populations
• HOWEVER, select harvest of mixed-conifer forests
  likely will markedly reduce habitat capability
  for flying squirrels
• Viability risk for Wrangell Island vole and POW
  flying squirrel likely less than suggested from
  research on these species elsewhere.

56
Information Needs

• ____________________________________
• Systematic inventory and genetic analysis of
  mammals, especially on nearshore and outer small
  (lt100,000 acre) islands
• Statistically robust sampling protocols to
  document absence of mammal taxa
• Long-term (gt5 years) population data
• Study vole and flying squirrel populations in
  managed stands
• Dispersal success of flying squirrels in managed
  landscapes.

57
Products from TLMP Support

• __________________________________________________
  ___________
• 2000. Foraging ranges of radio-marked marbled
  murrelets in southeast Alaska. Condor 102
  452-456.
• 2001. Dawn survey counts of marbled murrelets
  site and annual variation, sampling effort, and
  statistical power. Wildlife Society Bulletin
  29568-577.
• 1999. Relations of small mammal populations to
  even-aged shelterwood systems a comment.
  Journal of Wildlife Management, 63(4) 1376-1380.
  
• 1998. Bald eagle nesting in relation to clearcut
  logging in Southeast Alaska. Biological
  Conservation 83(2) 121-126.
• 1998. Increasing point count duration increases
  standard error. Journal of Field Ornithology,
  69(3) 450-456.
• 2001. Bird, mammal, and vegetation community
  surveys on Research Natural Areas in the Tongass
  National Forest. USDA Forest Service Research
  Paper-PNW-RP-535. Pacific Northwest Research
  Station Portland, OR. 44 p.
• 2002. Dietary uniqueness of northern flying
  squirrels in southeast Alaska. Canadian
  Field-Naturalist.

58
Products from TLMP Support (continued)

• ___________________________________________
• 2000. The northern flying squirrel (Glaucomys
  sabrinus) as a management indicator species for
  the Tongass National Forest Land and Resource
  Management Plan Assumptions, recent information,
  and priorities for studies and monitoring.
  Working Document, USDA Forest Service, Alaska
  Region, Juneau, AK.
• 2001. Small mammals and forest interactions
  mycorrhizal fungi as model organisms for
  understanding natural webs. Proceeding of the
  Non-timber Forest Products Convention, November
  2001, Anchorage, AK.
• In Press. Ecology and conservation of arboreal
  rodents of the Pacific Northwest. In Mammal
  community dynamics in western coniferous forests
  management and conservation, Zabel CJ, Anthony
  RG, editors. Cambridge University Press.
• In Press. Demography of the Prince of Wales
  Island flying squirrel an endemic of
  southeastern Alaska temperate rainforest.
  Journal of Mammalogy.
• In Review. Maintaining wildlife habitat in
  southeastern Alaska implications of new
  knowledge for forest management and research.
  (to Landscape and Urban Planning).

59
Products from TLMP Support (continued)

• ___________________________________________
• In Review. Demography of two endemic forest-floor
  mammals in southeastern Alaska temperate
  rainforest. Journal of Mammalogy.
• In Review. Sustainable management of wildlife
  habitat and risk of extinction Conservation
  Biology.
• In Review. Cost of transport in the northern
  flying squirrel, Glaucomys sabrinus. Journal of
  Mammalogy.
• In Review. Evolutionary diversity and ecology of
  endemic small mammals of southeastern Alaska with
  implications for forest management. (to Wildlife
  Monographs)
• In Review. Habitat correlates of flying squirrel
  abundance in temperate rainforests implications
  for ecosystem management. (to Ecological
  Applications)
• In preparation. Habitat correlates of abundance
  of two endemic forest floor mammals of
  southeastern Alaska temperate rainforests.