Partial migration in Oncorhynchus mykiss:

1
Partial migration in Oncorhynchus mykiss A
spatially and sexually explicit approach
J. McMillan photos
Justin Mills, USGS/OSU (MS, 2008) Jason Dunham,
USGS-FRESC Chris Jordan, NOAA-Fisheries Gordie
Reeves, USFS-PNW John McMillan, USGS/OSU (MS
2009) Chris Zimmerman, USGS
2
Sex and migration
Costs / benefits of migration Males Females
Decreased age-specific survival X X
Avoid poor freshwater conditions X X
Increased body size X X
Fitness strongly size dependent o X
J. McMillan photos
3
Space John Day River
4
Study objectives

1. Broad-scale measures of female anadromy
2. Predict patterns of female anadromy
3. Assess potential importance of local variability

5
Study design
Collect juvenile O. mykiss
Test for non-random distribution
Determine maternal origin
Collect water samples
Sites with anadromy
Tests of model performance
Broad-scale environmental variable(s)
Predictive model
Test for residual spatial variation
6
Collection and maternal origin
Four otoliths
Two fish water sample
P. Stratis photos
7
Anadromy was common, widespread
Offspring of
Steelhead 91
Rainbow trout 58
Anadromy at 52 of 72 sites
Two rainbow trout offspring
One of each
Two steelhead offspring
8
How is maternal origin distributed?
S
R
R
R
R
S
S
S
R
S
S
R
R
R
S
R
R
R
R
S
S
S
S
S
S
S
S
S
R
S
S
S
S
R
S
S
S
S
R
S
S
R
S
R
Random distribution
Numerical dominance or spatial segregation
9
Maternal origin was clustered
Combination at site Observed Expected
Different maternal origin 11 (23) 22.4 (48)
Same maternal origin 36 (77) 24.6 (52)
Both steelhead 23 (49) 17.3 (37)
Both rainbow trout 13 (28) 7.3 (15)
?² 11.15, df 1, P lt 0.001
n 47 sites only those with 2 juveniles lt 2
years old
10
Objective 2 Predictive model
Sites with anadromy
Tests of model performance
Broad-scale environmental variable(s)
Predictive model
Test for residual spatial variation
11
Stream size and anadromy

• Associated with many ecological and physical
  processes
• Sediment transport
• Water temperature
• Biological organization
• Readily used in spatial statistics
• Simple to estimate for large area

12
Anadromy varied with stream size
13
Mantel test for spatial autocorrelation
Euclidean distance
Stream network distance
14
No spatial autocorrelation

• Mantel tests non-significant
• Spatial gradients accounted for by model

Subset of 1/5 of pairwise distances
15
Bottom lines

• Sampling approach proved useful
• Female anadromy was predictable
• Stream size accounted for most of the broad-scale
  variability in female anadromy
• Local factors potentially source of remaining
  variability

J. McMillan photos
16
Improvements

• Model improvements
• Redd counts
• Combined probabilistic predictions
• Local factors
• Bioenergetics
• Species interactions
• Community effects
• Ecosystem processes
• Doesnt address males
• Doesnt address resident females

17
(No Transcript)
18
Discussion

• The process critical periods, sexual tension,
  and everything in-between
• The evidence observation, model, experiment
  correlation vs causation?
• The relevance ESA listing, modeling,
  monitoring, recovery du les sauvages?