1 Spatial and Spatial and Spatio-temporal Spatio-temporal modeling of the abundance of modeling of the abundance of spawning spawning c c oho oho salmon salmon on the on the Oregon coast Oregon coast R82-9096-01 Ruben Smith Don L. Stevens Jr. September 11, 2004
43
Embed
1 Spatial and Spatio-temporal modeling of the abundance of spawning coho salmon on the Oregon coast R82-9096-01 Ruben Smith Don L. Stevens Jr. September.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Spatial and Spatial and Spatio-temporal modeling of Spatio-temporal modeling of the abundance of spawning the abundance of spawning ccohooho salmon salmon
on the Oregon coaston the Oregon coast
R82-9096-01
Ruben Smith
Don L. Stevens Jr.
September 11, 2004
2
This presentation was supported under STAR Research Assistance Agreement No. CR82-9096-01 awarded by the U.S. Environmental Protection Agency to Oregon State University. It has not been formally reviewed by EPA. The views expressed in this document are solely those of authors and EPA does not endorse any products or commercial services mentioned in this presentation.
Coho SalmonCoho Salmon
3
OverviewOverview
• Introduction
• Part I : Spatial Analysis of the abundance of Coho salmon
• Part II: Spatio-temporal analysis of the abundance of Coho salmon
A male coho salmon with spawning colorationwww.zoology.ubc.ca/ ~keeley/coho4.jpg
4
IntroductionIntroduction
• Coho salmon spend their adult lives at sea and return to natal streams along the Oregon Coast to spawn
• In 1960s and early 1970s Coho salmon were easily available for fishing off Oregon coast
• By late 1970s there were signs that Coho salmon stocks have declined in some regions of their range
• Coho salmon in Oregon coastal basins are listed as threatened under the Endangered Species Act
A male coho salmon with spawning colorationwww.zoology.ubc.ca/ ~keeley/coho4.jpg
5
IntroductionIntroduction
• The Oregon Department of Fisheries and Wildlife (ODFW) divides the coastal streams in four monitoring areas:
North Coast
Mid-Coast
Mid-South Coast,
Umpqua based on genetic variation and life-history traits
6
IntroductionIntroduction
• Sites are selected using a rotating annual panel sample design (Stevens, 1997; Stevens & Olsen, 2000, 2002)
• The sampling unit is about 1-mile long stream reach (site)
A male coho salmon with spawning colorationwww.zoology.ubc.ca/ ~keeley/coho4.jpg
7
DataData• ODFW winter spawning
Coho surveys
• Visual counts of spawning Coho is used in the stream surveys
• Number of surveyed sites:
~ 420 sites per year.
~105 per monitoring areas
• Data available for years : 1998-2003 http://www.surfingvancouverisland.com/fish/images/st00b01.jpg
8
Coho Population UnitsCoho Population Units
• The ODFW identified 33 Coho salmon populations units on 4 monitoring areas (North Coast, Mid-Coast,Mid-South Coast and Umpqua)
based on:
• geography
• similarity of habitats
• extinction risk
• potential similarity of life history types
North CoastNorth Coast
Mid-CoastMid-Coast
Mid-South Mid-South
CoastCoastUmpquaUmpqua
Coho Population UnitsCoho Population Units
9
Spawner abundance has generally been lower in the south/north and higher in the mid-coast of Oregon
Max=270 Max=326 Max=1659
Radius of circle are prop. to observed count
+ zero counts
10
Spatial Analysis Goal
To generate prediction maps of abundance of
spawners for each year on the Oregon Coast
11
• Diggle et.al. (1998)
• We fitted a separate model for each year (1998-2003)
• Notation
: observed count of adult spawners
at Coho population , 1, ,
and spatial sampling site , 1, ,
ji
ji j
y
j j P
i n
s
Spatial Analysis Spatial Analysis --Individual Year Analysis
Poisson Generalized Linear Geostatistical ModelPoisson Generalized Linear Geostatistical Model
12
• Model for the dataModel for the data:
Yji| λji ~ independent Poisson(lji λji)
• Model for the log of the Poisson rateModel for the log of the Poisson rate
Spatial Analysis Spatial Analysis - - Individual Year Analysis
Poisson Generalized Linear Geostatistical ModelPoisson Generalized Linear Geostatistical Model
non-spatial random effect
length of the site ssjjii (kms)
density of spawning Coho at site s sjji i (counts per km)
spatially correlated random effect
log( )ji j ji jiZ
Dependence among observations is induced by the random spatial process ji
13
Spatial Analysis (cont.)Spatial Analysis (cont.)
• Assume
is a correlation parameter
Sites from different coho populations are assumed
independent
21, , MVN( , ) 1, ,
jj n z jZ Z j P Z 0 R = ~
2 ~ MVN ( , ) 1, ,jj n j P η 0 I
( , ) Corr ,
( , ) exp with ( , ) || ||
ji ji ji ji
jj ji ji
Z Z Z Z
d i id i i
R
s s
14
Model FittingModel Fitting
• We placed prior on the parameters and compute the joint posterior distribution given the data
where
2 2 2
2 2 2
, , , , | | , ,
| , ( ) ( ) ( )
z
z z
p p p
p
λ Z μ y y λ λ Z μ
Z μ
, |
exp λ μ Z η
15
• The likelihood is analytically intractable
• We use Markov Chain Monte Carlo (MCMC) methods to estimate the parameters
Gibbs sampler
Metropolis-Hastings algorithm
• A MATLAB computer program was used to simulate realizations from the posterior distributions of , z
2, , 2
and each of the elements of Z, and λ to generate a Markov chain.
Individual Year Analysis (cont.)Individual Year Analysis (cont.)
16
PredictionPrediction
• To obtain predictions at the prediction grid locations of
the spatial component, ZZp
the density of returning
adult coho, p
we sample from their complete conditional distributions:
• p(ZZp |Z, , , zz22)
• p(p |ZZp, , , 22 )
Prediction Grid
Center of prediction grid box denoted by +
17
18
GoalsGoals
• Predict the spatial abundance of Returning adult Coho over time.
• For this presentation we considered only 18 Coho populations.
A male coho salmon with spawning colorationwww.zoology.ubc.ca/ ~keeley/coho4.jpg
Spatio-Temporal ModelSpatio-Temporal Model
19
Radius of circle are prop. to observed count
+ zero counts
Counts of returning adult Coho salmonCounts of returning adult Coho salmon1998-20031998-2003
20
Yearly Counts by Coho Population (18)Yearly Counts by Coho Population (18)
21
• Wikle (2003)
• Notation
• Concern about the short time series available
: observed count of spawning Coho salmon at year ,
coho population and site
for 1,..., (1998, , 2003);
1,..., (1, ,18);
1,...,
tji
ji
tj
y t
j
t T
j P
i n
s
Spatio-Temporal Analysis Spatio-Temporal Analysis Poisson Generalized Linear Geostatistical ModelPoisson Generalized Linear Geostatistical Model
22
Spatio-Temporal ModelSpatio-Temporal ModelData ModelData Model
• Model for the data:
Ytij| λtij ~ independent Poisson(ltjiλtij)
• Dependence among observations is induced by the random spatio-temporal process
density of spawning coho at site at time tjis
tji
length of the site ssjji (kms) at time t
23
Process ModelProcess Model
log tj i tji tj tji k Z
known vector that relates sampled locations with the Z-process.
Each sampled site is assigned to the nearest grid location
spatio-temporal random process that accounts for observational error and small-scale spatio-temporal variation
Spatio-temporal dynamic process that accounts for the spatial variation of the coho spawners over time
is an m1 vector representation of the gridded Z-process at the prediction locations