Estimating Greater Sage-grouse Juvenile Survival in Utah

1
Estimating Greater Sage-grouse Juvenile Survival
in Utah

• Utah State University
• David Dahlgren
• Terry Messmer
• David Koons

2
Introduction

• Info Need

3
Study Area

• Parker Mountain

4
Methods
5
Methods Burkepile et al. 2002
6
Methods

• Hen Behavior

7
Methods

• Monitoring

8
Methods

• Veg and Arthropod Sampling

9
Brood Mixing
10
Methods

• Cause of mortality
• Handling
• Exposure
• Predation
• Avian/Mammalian
• Unknown

11
Methods

• Modeling
• Manly and Schmutz 2001 - JWM
• Maximum Likelihood Estimation (Version of the
  Mayfield Estimator)
• Heterogeneity - D

Brood ID No. chicksstart No. chicks end Age (days) Start Age (days) End Covariates
807 6 5 2 3 Year etc.
807 5 5 3 6 Year etc.
807 5 5 6 9 Year etc.
807 5 5 9 11 Year etc.
807 5 5 11 13 Year etc.
807 5 5 13 15 Year etc.
807 5 4 15 17 Year etc.
12
Methods

• Modeling Survival
• First
• Age Structure (weeks 1 to 6)
• Used to assess covariates
• AIC
• Second
• Assess covariates
• Temporal
• Brood hen characteristics
• Vegetation data
• Arthropod data

13
Methods

• Modeling Survival
• Assumptions
• Brood-mixing and right censoring
• Missing chicks
• Brood-mixing and missing chicks

14
Methods

• Modeling Survival
• First Covariates
• Year
• Brood Type
• Hen Behavior (restricted data set)
• Hen Age (restricted data set)
• Hatch Date

15
Methods

• Modeling Survival
• Second Vegetation Covariates (restricted data
  set)
• Shrub cover and height
• Grass cover and height
• Forb cover and height
• Third Arthropod Covariates (2006, restricted
  data set)
• Hymenoptera (ants separate), Coleoptera,
  Lepidoptera, Orthoptera, miscellaneous, and total
  arthropods

16
Results

• Sample sizes
• Total
• 2005 n 89 chicks in 21 broods (mean 4.24)
• 2006 n 61 chicks in 21 broods (mean 2.91)
• Handling Death (2.6)
• 2005 n 3
• 2006 n 1
• Excluded from the survival analysis

17
Results

• Sample Sizes
• Chick mass
• Mean 29.5g (SE 0.16)
• Transmitter averaged 5.1 (SE 0.0003) of chick
  weight
• Hen Behavior
• 43 Very protective (18/42)
• 38 Moderately protective (16/42)
• 19 Non-protective (8/42)

18
Results

• Sample Sizes
• Brood Mixing
• 21 of marked chicks (31/146)
• 43 of marked broods (18/42)
• 45 of mixing events involved gt1 chick (9/20)
• Occurred weeks 1 to 6
• 70 (14/20) in weeks 2 and 3
• Radio-marked hen mortality (n 2)

19
Results

• Sample Sizes
• Chick Mortality
• n 44 documented deaths
• n 26 missing (assumed depredated)
• n 6 exposure
• Predation 91 (64/70)
• Unknown 75 (48/64)
• Mammalian 12.5 (8/64)
• Avian 12.5 (8/64)

20
Results
Model K AIC wi
Age Specific Models (no covariates)
age (week1)(week2)(week3)(week4)(weeks5-6) 6 345.89 0.00 0.99973
age (weeks1-2)(weeks3-4)(weeks5-6) 4 362.47 16.58 0.00025
age (week1)(week2)(week3)(weeks4-6) 5 368.95 23.06 0.00000
age (weeks1-2)(weeks3-6) 3 372.67 26.78 0.00000
age (week1)(week2)(weeks3-6) 4 374.64 28.75 0.00000
age (week1)(weeks2-6) 3 398.40 52.51 0.00000
age (weeks1-2)(weeks4-6) 3 400.48 54.59 0.00000
age (weeks 1-6) 2 408.52 62.63 0.00000
Covariate Models
age brood type (regular or mixed) 7 253.60 0.00 0.99999
age hen age (yearling or adult) 7 279.42 25.82 0.00003
age year (2005 or 2006) 7 332.26 78.66 0.00000
age hen behavior (protectiveness) 8 335.07 81.47 0.00000
age hatch date (Julian days) 7 343.07 89.47 0.00000
First Age parameterization
Second Temporal and hen characteristics
AIC difference between a model (i.e., model i)
and the best performing model (i.e., model with
the lowest AIC among the set of models
examined). wi Akaike model weight. The best
model of age (week1) (week2) (week3)
(week4) (weeks5-6)
21
Results
First Age parameterization
Model K AIC wi
Age Specific Models (no covariates)
age (week1)(week2)(week3)(week4)(weeks5-6) 6 345.89 0.00 0.99973
22
Results
Second Temporal and hen characteristics
Model K AIC wi
Covariate Models
age brood type (regular or mixed) 7 253.60 0.00 0.99999
23
Results
Mean Survival to 42 days 0.41 (SE
0.046) Chick Survival in Regular broods 0.38
Chick Survival in Mixed broods 0.61
24
Results

• Heterogeneity (D)
• Chicks/brood 3.5
• For best model
• D 1.10 (SE 0.22)

25
Results

• Vegetation Covariates
• Null Model is best

Model K AIC wi
age (NULL) 6 -19.48 0.00 0.99999
26
Results

• Arthropod Covariates
• Entire 42 days
• Ant model

Model K AIC wi
age Ants 7 -115.16 0.00 0.59508
age (NULL) 6 -114.39 0.77 0.40492
27
Results

• Arthropod Covariates
• Early brood-rearing (day 1-21)
• Orthoptera (grasshopper) model
• Estimates not significant

Model K AIC wi
age Orthoptera 5 -24.03 0.00 0.99999
28
Take Home

• Predation major cause of chick mortality
• However, survival was good (mean 0.41)
• Our data suggested low dependence among brood
  mates for sage-grouse chicks
• Brood-mixing may be important to survival, needs
  further investigation
• There is evidence that Arthropods (especially
  Orthoptera) may aid chick survival, needs further
  investigation (gt sample size)

29
Thanks

• PARM
• USU Extension
• Parker Mtn. Grazing Association
• Jack H. Berryman Institute
• UDWR
• USFS
• BLM
• NRCS
• SITLA
• Farm Bureau
• County Commissioners

• People
• Technicians
• Nathan Burkepile
• Jack Connelly
• Volunteers

30
Any Questions?