ANATOMY OF SAMPLING STUDIES OF ECOLOGICAL RESPONSES THROUGH TIME {on the web at http://www.oregonstate.edu/instruct/st571/urquhart/anatomy/index.htm} by N.
Download ReportTranscript ANATOMY OF SAMPLING STUDIES OF ECOLOGICAL RESPONSES THROUGH TIME {on the web at http://www.oregonstate.edu/instruct/st571/urquhart/anatomy/index.htm} by N.
Slide 1
ANATOMY OF SAMPLING STUDIES
OF
ECOLOGICAL RESPONSES THROUGH TIME
{on the web at http://www.oregonstate.edu/instruct/st571/urquhart/anatomy/index.htm}
by
N. Scott Urquhart
Oregon State University, USA
and
Anthony R. Olsen
US EPA
MSTS/2004
ANATOMY # 1
Slide 2
THE AUTHORS
N. SCOTT URQUHART
Trained in Statistics
About 40 Years of Experience in Applications
Worked With Ecologists in Desert, Arctic, Pacific Northwest
Many Surveys with Rural Sociologists and Ag Economists
Including 10 years with EPA’s Environmental Monitoring and
Assessment Program (EMAP)
ACADEMIC And AGENCY; PLANT And ANIMAL
ANTHONY (Tony) R. OLSEN
Trained in Statistics
30+ Years of Experience in Private and Government Applications
Worked With Atmospheric Modelers And Air Pollution Field
Scientists
Survey Experience With Health Professionals And Large-scale
National Resource Monitoring
Now Statistical Lead with EPA’s EMAP
MSTS/2004
ANATOMY # 2
Slide 3
EVOLUTION OF THE “ANATOMY”
The first step in the development of the
ANATOMY focused on experimental design
situations.
Served as the structure for several part-semester
courses in advanced statistical methods at New
Mexico State University
Eventually published as
Urquhart, N. S. (1981). Anatomy of a study.
HortScience 16:621-627.
Experience with EMAP led to its expansion to
surveys
MSTS/2004
ANATOMY # 3
Slide 4
TODAY’S CONTEXT for
SURVEYS
“EMAP-type Situations”
EMAP = US EPA’S Environmental
Monitoring and Assessment Program
Estimate Status, Changes ...
In Indicators
Estimate Status, Changes, ...
In Extent
Describe Associations ...
MSTS/2004
ANATOMY # 4
Slide 5
Objective #1: Estimate the status, changes and trends in
selected indicators of the condition of our Nation's
ecological resources on a regional scale
with known confidence
17.6%
± 10%
6.8%
± 6%
31.8%
± 8%
43.8%
± 12%
Hypereutrophic
(N=258)
Eutrophic
Mesotrophic
Oligotrophic
Source: EMAP Northeast
MSTS/2004
Lakes Study
ANATOMY # 5
Slide 6
Objective #2: Estimate the status, changes and trends
in the extent and geographic coverage of our Nation's
ecological resources on a regional scale with known
confidence
12000
Est.
Lake #
North
east
SE
Est.
Area
SE
10000
8000
11,455
1,251
4,030
814
6000
Adir
1,506
285
1,082
395
4000
NEU
C/L/P
5,689
4,280
1,206
1,048
2,099
758
850
254
2000
0
Northeast
Adir = Adirondacks;
NEU = New England Uplands;
MSTS/2004
C/L/P
= Coast & Lake Plains
Adir
NEU
C/L/P
Source: EMAP NortheastANATOMY
Lakes Study
#6
Slide 7
Objective #3: Describe associations
between indicators of anthropogenic stress
and indicators of condition
Fish Index of Biotic Integrity
Relative Ranking of Stressors
Good
Sedimentation
(Insufficient
Data)
17%
Riparian Habitat
24%
Mine Drainage
17%
31%
11%
10%
Tissue Contamination
Fair
Phosphorus
5%
Nitrogen
5%
Acid Mine Drainage
0%
Proportion of Stream Length
Source: EMAP Mid-Atlantic
MSTS/2004
14%
Acidic Deposition
36%
Poor
25%
Highlands Assessment
1%
10%
20%
30%
40%
% of Stream Length
ANATOMY # 7
Slide 8
WHO MUST COMMUNICATE
MSTS/2004
Ecologists & Other Biologists
Statisticians
Geographers
Geographic Information Specialists (GIS)
Information Managers
Quality Assurance Personnel
Managers, At Various Levels
ANATOMY # 8
Slide 9
“SAMPLING”
A WORD OF MANY MEANINGS
Statisticians Often Associate It With Survey
Sampling
An Ecologist May Associate It with the Selection of
Local Sites or Material
A Laboratory Scientist May Associate It With the
Selection Of Material to be Analyzed from
Material Supplied
Common General Meaning, Varied Specific
Meanings
MSTS/2004
ANATOMY # 9
Slide 10
THE SPECIAL NEED
Communication Demands a Distinction
Between
The
Local Process of Evaluating a Response,
and
The Statistical Selection of a Sampling Unit,
For example,
A LAKE
A POINT ON A STEAM
MSTS/2004
ANATOMY # 10
Slide 11
THE SPECIAL NEED - continued
The Terms
Response Design
Sampling Design or Survey Design
Can Be Used to Make this Distinction
But a Complex Ecological Survey
Clearly Has More Parts Than These!
MSTS/2004
ANATOMY # 11
Slide 12
BASIC ROLES
Survey Design Tells Us Where to Go to
Collect Sample Information or Material
Response Design Tells Us What to Do
Once We Get There
But These Two Components Exist in a
Broader Context
MSTS/2004
ANATOMY # 12
Slide 13
AN IMPORTANT DISTINCTION
Monitoring Strategy
Conceptual
Impacted by Objectives
Addressable Without Regard to
the Inference Strategy
Inference Strategy
MSTS/2004
ANATOMY # 13
Slide 14
AN IMPORTANT DISTINCTION
- continued
Monitoring Strategy
..........
Inference Strategy
Places to Evaluate the Response – “the
WHERE”
Relation Between Points Evaluated and
the Population
IE, the Basis for Inference
MSTS/2004
ANATOMY # 14
Slide 15
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
Monitoring Strategy
Universe Model
Statistical Population
Domain Design
Response Design
These components
exist regardless
of the
inference strategy
Inference Strategy
Survey Design
Temporal Design
These components
exist for any
monitoring strategy
Quality Assurance Design
MSTS/2004
ANATOMY # 15
Slide 16
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 16
Slide 17
The MONITORING STRATEGY
The MONITORING STRATEGY MUST
RESPOND TO
Monitoring Objectives
State of Knowledge in Ecological Sciences
Characteristics of Ecological Resource(s)
of Interest
EXPECTED FUNDING Compared To COSTS
Operational Constraints
MSTS/2004
ANATOMY # 17
Slide 18
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 18
Slide 19
The UNIVERSE MODEL
Reality (Universe): Ecological Entity Within
a Defined Geographic Area to Be Monitored
Model of the Universe:
Development of a Monitoring Approach
Requires Construction of a Model for the
Universe
Elements Of The Universe Model: Set of
Entities Composing The Entire Universe
MSTS/2004
ANATOMY # 19
Slide 20
The UNIVERSE MODEL
Population Description and Its Sampling
Require Definition of the “Units” in the
Population
Discrete Units:
Lakes May Be Viewed This Way
Continuous Structure in Space of Some
Dimension:
2-space: forests or agroecosystems
1-space: Streams
3-space: Ground Water
MSTS/2004
ANATOMY # 20
Slide 21
THE MODEL FOR STREAMS
Strahler Orders
Second
Order
First Orders
First Orders
First Orders
First Order
MSTS/2004
ANATOMY # 21
Slide 22
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 22
Slide 23
The STATISTICAL POPULATION
The Collection of Units (as modeled) Over Some
Region of Definition
Spatial
Temporal
SPATIAL And TEMPORAL
Population Definition Could Include Features
Which Depend on Response Values
EX: acid sensitive streams at upper elevations
MSTS/2004
ANATOMY # 23
Slide 24
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 24
Slide 25
The DOMAIN Design
Specifies Subpopulations or “Domains” of
Special Interest
May Specify Meaningful Comparisons
Between Domains
Similar to Planned Comparisons in
Experimental Design Situations
Domain Design May Depend on Response
Values
Ex: Warm Versus Cold Water Lakes
MSTS/2004
ANATOMY # 25
Slide 26
The DOMAIN DESIGN - continued
Specifies Subpopulations or
“Domains” of
Special Interest
Determined From Defining Factors For
The Monitoring Activity
Must Have Critical Connection To Clients
Other Domains May Be Used For Analysis,
Without Having Been Used In Defining The
Monitoring Strategy
EX: EMAP domains include ECOAREAS
and STANDARD FEDERAL REGIONS
MSTS/2004
ANATOMY # 26
Slide 27
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 27
Slide 28
The RESPONSE Design
The Response Design Specifies
The Process Of Obtaining A Response
At An Individual Element (Site)
Of The Resource
During A Single Monitoring Period
Response: What Will Be Determined
On An Element –
Needs To Be Responsive to the Objectives of the
Monitoring Activity
MSTS/2004
ANATOMY # 28
Slide 29
The RESPONSE Design - continued
EMAP Responses Focus On Indicators of
STRESS and
Condition
The Response Design Also Defines
Plot Design
Measurement Protocols
Support Region – area around the site where material
is collected, or measurements are taken
Data Reduction Protocols
Calculation Of The Final Indicator Value for the
Element
MSTS/2004
ANATOMY # 29
Slide 30
The RESPONSE Design
- Continued
For example, consider a response related to
macroinvertebrates in streams
MSTS/2004
RESPONSE = proportion EPT (This is the
proportion of collected macrobenthos
organisms, mainly insects, which fall in the
taxonomic classes of Ephemeroptera ,
Plecoptera , or Trichoptera. Low values
indicate polluted streams; high values indicate
rather pristine streams)
ANATOMY # 30
Slide 31
The RESPONSE Design - continued - 2
... response related to macrobenthos ...
The COLLECTION UNITS could be 10
30cm x 30cm areas, systematically organized, at
the stream site, sampled with a “Surber
sampler”
The EVALUATION UNIT could be a jar
containing the composite of all macroinvertebrate
organisms collected at the 10 collection sites, or
MSTS/2004
The EVALUATION UNIT also could be a jar
containing a 1/6 subsample of the composite of
macroinvertebrate organisms collected in the
10 collection units.
ANATOMY # 31
Slide 32
The RESPONSE Design - continued - 3
... response related to macrobenthos ...
The LABORATORY EVALUATION of the
material would consist of determining and
recording the taxa (like family, genus, or
species) of each organism in the evaluation
material
The RESPONSE would be determined by
computing the number of organisms in the
evaluation material belonging to the E, P, T
taxonomic classes, and dividing this by the
number of organisms classified.
ANATOMY # 32
MSTS/2004
Slide 33
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 33
Slide 34
The INFERENCE STRATEGY
Is The Basis For Scientific Inference
Provides The Connection Between Objectives and the
Monitoring Strategy
Monitoring Strategy Usually Must Rely on Obtaining
Information on a Subset Of All Possible Elements in
the Universe
Specifies Which Elements of the Universe Will Have
Responses Determined on Them
Can Be Based On Either ... (continued )
MSTS/2004
ANATOMY # 34
Slide 35
The INFERENCE STRATEGY
(continued)
... Connection ...
... Subset ...
... Have Responses
Can Be Based On Either
Judgment Selection Of Units
Inferential Validity Rests on Knowledge Of Relation
Between the Universe And the Units Evaluated
Probability Selection Of Units
(The Focus Here)
MSTS/2004
ANATOMY # 35
Slide 36
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 36
Slide 37
The SURVEY Design
Probability Based Survey Designs Are
Considered Here
May Be Somewhat Limited To Sedentary Resources
Positive Features
(As An Observational Study)
MSTS/2004
Permit Clear Statistical Inference to
Well-Defined Populations
Measurements Often can be Made in Natural
Settings, Giving Rise to Greater Realism
Eventual Results
ANATOMY # 37
Slide 38
The SURVEY DESIGN - CONTINUED
Disadvantages
Limited Control Over Values of Predictor
Variables
Restricts Causative Inference
Usually Will Produce Inaccessible Sampling
Points
Good - For Inference
Bad - For Logistics
MSTS/2004
ANATOMY # 38
Slide 39
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
MSTS/2004
ANATOMY # 39
Slide 40
VIEW DOWN TRANSECT AT
MILE 12.3
MSTS/2004
ANATOMY # 40
Slide 41
CLIFF AT
MILE 135.2
(PARTIAL
HEIGHT)
MSTS/2004
ANATOMY # 41
Slide 42
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
Example: Frame for selecting field sites on streams
in the Western US
MSTS/2004
ANATOMY # 42
Slide 43
FRAME ERRORS
TO BE DOCUMENTED SHORTLY
Water Body Size
Flow Status -- re Perennial
Identified As Perennial, but not correct
Wastes Effort Of Field Crews
Identified as Non-perennial, but Really is
Perennial
Missed Resource
Inaccurate Assessment
MSTS/2004
ANATOMY # 43
Slide 44
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 44
Slide 45
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 45
Slide 46
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 46
Slide 47
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
Example: Frame for selecting field sites on streams
in the Western US
May change over time
– As, for example, land use changes
MSTS/2004
ANATOMY # 47
Slide 48
SITE SELECTION
Needs to Accommodate Realities Such As
Frame Imperfection
Frame Which Changes Over Time
Sites nearly Uniform Over the Resource
But with substantial randomization
Supports Variable Probability of Selection
Generalized Randomized Tessellation Stratified
Sampling = GRTS
The topic of the next session
MSTS/2004
ANATOMY # 48
Slide 49
The EMAP SURVEY Design
Assures Representation and Inference to
Populations
Adapted to Resource Characteristics
Emphasizes Spatial Allocation of Samples
Uses Two-phase Sampling; Phase I
Based on a Randomized Point Grid and
Associated Areas
MSTS/2004
ANATOMY # 49
Slide 50
EMAP EXAMPLE OF SELECTED
SITES
Mid-Appalachian Highlands Stream Pilot
MSTS/2004
ANATOMY # 50
Slide 51
MSTS/2004
ANATOMY # 51
Slide 52
EMAP EXAMPLE OF SELECTED
SITES
Western Stream Pilot
Non-perennial/perennial survey
Survey of perennial streams
MSTS/2004
ANATOMY # 52
Slide 53
MSTS/2004
ANATOMY # 53
Slide 54
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 54
Slide 55
The TEMPORAL Design
The TEMPORAL DESIGN specifies the
pattern of revisits to sites selected by the
Survey Design
Sampled population units are partitioned into
one (degenerate case) or more PANELS.
Each population unit in the same panel has the
same temporal pattern of revisits.
Panel definition could be probabilistic or
systematic
Several temporal designs follow
MSTS/2004
ANATOMY # 55
Slide 56
TEMPORAL DESIGN:
ROTATING PANEL
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1
2
3
4
5
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2
3
4
5
6
7
8
9
MSTS/2004
6
7
8
9
1 0 1 1 1 2 1 3 ...
X
X
ANATOMY # 56
Slide 57
TEMPORAL DESIGN:
ROTATING PANEL
A ROTATING PANEL DESIGN IS THE TEMPORAL
DESIGN USED BY THE NATIONAL AGRICULTURAL
STATISTICAL SERVICE (US - “NASS”)
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE
EXPERIMENTAL DESIGN SENSE
IT IS FAIRLY WELL SUITED FOR ESTIMATION
“STATUS,” BUT NOT NEARLY PARTICULARLY
POWERFUL FOR DETECTING TREND OVER
INTERMEDIATE TIME SPANS
MSTS/2004
ANATOMY # 57
Slide 58
TEMPORAL DESIGN:
SERIALLY ALTERNATING
T IM E PE R IO D ( ex: Y E A R S )
PA N E L
1
1
X
2
3
4
2
3
4
5
6
7
8
X
X
X
X
X
X
10 11 12 13 ...
X
X
X
9
X
X
X
THIS TEMPORAL DESIGN IS “UNCONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE. THIS
WAS
THE TEMPORAL DESIGN INITIALLY PRESCRIBED
FOR EMAP, BUT ...
MSTS/2004
ANATOMY # 58
Slide 59
TEMPORAL DESIGN:
AUGMENTED SERIALLY ALTERNATING
T IM E P E R IO D ( ex: Y E A R S )
PA N E L
1
2
3
4
5
6
7
8
9
0
X
X
X
X
X
X
X
X
X X
1
X
2
3
4
X
X
X
X
X
X
X
X
X
X
X
X
10 11 12 13 ...
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 59
Slide 60
TEMPORAL DESIGN:
PARTIALLY AUGMENTED
SERIALLY ALTERNATING
T IM E P E R IO D ( ex: Y E A R S )
PA N E L
1
2
3
4
5
0
X
X
X
X
X
1
X
2
3
4
6
7
8
X
X
10 11 12 13 ...
X
X
X
9
X
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 60
Slide 61
TEMPORAL DESIGN:
SERIALLY ALTERNATING
WITH CONSECUTIVE YEAR REVISITS
T IM E PE R IO D ( ex: Y E A R S )
PA N E L
1
2
1
X
X
2
3
4
X
3
4
5
6
X
X
X
X
X
X
X X
7
8
9
10 11 12 13 ...
X
X
X
X
X
X
X
X
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE
EXPERIMENTAL DESIGN SENSE
BUT IT PROVIDES VISITS TO ONLY HALF AS MANY
SITES AS THE BASE SERIALLY ALTERNATING DESIGN
MSTS/2004
ANATOMY # 61
Slide 62
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH CONSECUTIVE YEAR
REVISITS
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1
1
X
1A
X
2
3
4
5
6
7
8
X
X
2
X
2A
X
X
9
X
X
X
X
X
3
X
3A
X
X
4
X
4A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1 0 1 1 1 2 1 3 ...
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 62
Slide 63
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
T IM E P E R IO D ( ex: Y E A R S )
PANEL
1
1
X
1A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
MSTS/2004
2
3
4
5
6
7
8
X
9
10
11
12
X
X
13
...
X
2
X
X
2A
X
X
X
X
X
X
X
X
X
X
X
X
3
X
X
X
3A
X
X
X
X
X
X
X
X
X
X
X
X
4
X
X
X
4A
X
X
X
X
X
X
X
X
X
X
X
X
ANATOMY # 63
Slide 64
THE REVISIT SPLIT PANEL OF THE
THE TEMPORAL DESIGN IN USE BY
EMAP - SURFACE WATERS
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1A
2A
3A
4A
MSTS/2004
1
2
X
X
X
X
X
X
3
4
5
6
X
X
X
X
X
X
X
X
X
X
X
X
X
X
7
9
X
X
X
X
X
X
X
X
8
1 0 1 1 1 2 1 3 ...
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
ANATOMY # 64
Slide 65
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
THIS TEMPORAL DESIGN IS “CONNECTED” IN
THE EXPERIMENTAL DESIGN SENSE
IT ALSO SUPPORTS ESTIMATION OF THE
SITE by TIME PERIOD (SITE by YEAR)
INTERACTION
REVISITS TO ABOUT 10% OF SITES ALLOCATES
ABOUT 30% OF RESOURCES TO REVISITS
MSTS/2004
ANATOMY # 65
Slide 66
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
THIS TEMPORAL DESIGN IS “CONNECTED” IN
THE EXPERIMENTAL DESIGN SENSE
IT ALSO SUPPORTS ESTIMATION OF THE SITE by
TIME PERIOD (SITE by YEAR) INTERACTION
WE WILL RETURN TO THIS MATTER AFTER WE
CONSIDER THE QUALITY ASSURANCE DESIGN.
REVISITS TO ABOUT 10% OF SITES ALLOCATES
ABOUT 30% OF RESOURCES TO REVISITS
MSTS/2004
ANATOMY # 66
Slide 67
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 67
Slide 68
QUALITY ASSURANCE DESIGN
Defines Those Activities Intended to Provide Data
of Known Quality:
Blind duplicates
Accepted chemical standards, Etc
Can Provide Valid Estimates of the Variance of
Pure Measurement Error
MSTS/2004
ANATOMY # 68
Slide 69
RELEVANT COMPONENTS OF
VARIANCE FOR EMAP
POPULATION = LAKE or STREAM, for example
YEAR YEAR by SITE CREW SHORT TERM TEMPORAL - index window PROTOCOL ERROR -
MEASUREMENT ERROR -
MSTS/2004
ANATOMY # 69
Slide 70
RELEVANT COMPONENTS OF
VARIANCE FOR EMAP
WHAT DOES IT TAKE TO ESTIMATE THESE?
WHICH ONES SHOULD BE INCLUDED IN
“RESIDUAL” VARIANCE
WHICH TREND MUST OVERCOME TO BE
DEMONSTRATED?
MSTS/2004
ANATOMY # 70
Slide 71
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
UNIVERSE MODEL
(AS DISTINCT FROM THE UNIVERSE)
NOT RELEVANT
VIEWED AS A LIST
POPULATION
MSTS/2004
VERY SIMILAR, OTHER THAN DIFFERENCES
IMPLICIT IN THE UNIVERSE MODEL
ANATOMY # 71
Slide 72
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued
DOMAIN DESIGN
USUALLY NOT EXPLICITLY ACKNOWLEDGED,
BUT VIEWED AS PART OF THE DESIGN
PROCESS
RESPONSE DESIGN
QUESTIONNAIRE CONSTRUCTION AND DESIGN
SOME ELEMENTS OF INTERVIEW PROTOCOLS
SURVEY DESIGN
MSTS/2004
RELATIVELY SIMILAR
ANATOMY # 72
Slide 73
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued - 2
TEMPORAL DESIGN
PRESENT, BUT
MANY FINITE POPULATION SURVEYS ARE ONE-TIME
MANY LARGE SURVEYS HAVE A TEMPORAL
DIMENSION
CHANGE USUALLY IS OF FAR MORE INTEREST
THAN TREND
SURVEY LITERATURE DISTINGUISHES
BETWEEN
GROSS CHANGE - follows units across time
NET CHANGE - recognizes that change can occur
MSTS/2004
ANATOMY # 73
in
Slide 74
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued - 3
QUALITY ASSURANCE DESIGN
SURVEYS OF HUMANS (OR BUSINESSES, ETC)
HAVE A VERY SIMILAR INTENT, BUT DETAILS
DIFFER SUBSTANTIALLY
SUPERVISION IN PHONE SURVEYS
QUESTIONS TO CONFIRM RESPONDENT
CONSISTENCY
REINTERVIEWS and SIMILAR
REEVALUATIONS
MSTS/2004
ANATOMY # 74
Slide 75
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
UNIVERSE MODEL
STATISTICAL POPULATION
DOMAIN DESIGN
RESPONSE DESIGN
These components
exist regardless
of the
inference strategy
INFERENCE STRATEGY
SURVEY DESIGN
TEMPORAL DESIGN
These components
exist for any
monitoring strategy
QUALITY ASSURANCE DESIGN
MSTS/2004
ANATOMY # 75
Slide 76
LINKS TO MONITORING
REPORTS/DOCUMENTATION
Links to various monitoring programs,
To the program’s site
One to its methods, and
One to a current report.
As links change in an unpredictable fashion,
The “search words” should provide a quick path to the
current link.
Speaker on this program representing that program
MSTS/2004
ANATOMY # 76
Slide 77
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Adelaide Coastal Waters Study
Program site:
http://www.environment.sa.gov.au/epa/acws.html
http://www.clw.csiro.au/acws/ (more technical of the two)
Current Report:
http://www.environment.sa.gov.au/epa/pdfs/acwsnewsno5.pdf
Methods:
http://www.clw.csiro.au/acws/IS1.html
Search words: adelaide coastal waters study; acws au
Speaker: David Fox, University of Adelaide
MSTS/2004
ANATOMY # 77
Slide 78
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Alberta Biodiversity Monitoring Program (ABMP)
Program site:
http://www.abmp.arc.ab.ca/
Current Report:
http://www.abmp.arc.ab.ca/AnnualReport2003.pd
f
Methods:
http://www.abmp.arc.ab.ca/ScienceProtocols.htm
Search words: alberta biodiversity monitoring
program; abmp ca
Speaker: None, unfortunately
MSTS/2004
ANATOMY # 78
Slide 79
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Environmental Monitoring and Assessment
Program (EMAP)
Program site:
http://www.epa.gov/emap/
Current Report:
http://www.epa.gov/maia/html/maha.html
Methods:
http://www.epa.gov/emap/html/pubs/docs/groupdocs/
surfwatr/field/ws_abs.html
Search words: epa environmental monitoring assessment
program; emap epa
Speaker here: Tony Olsen
MSTS/2004
ANATOMY # 79
Slide 80
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Forest Inventory and Analysis (FIA)
Program site:
http://fia.fs.fed.us/
Current Report:
http://www.fs.fed.us/research/sustain/
(for example)
Methods:
http://fia.fs.fed.us/FIAProgramInformation.htm
(accessible at this link)
Search words: forest inventory analysis; FIA USDA
Speaker here: Mike Williams
MSTS/2004
ANATOMY # 80
Slide 81
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Agricultural Statistical Service
(NASS)
Program site:
http://www.usda.gov/nass/
Current Reports:
http://www.usda.gov/nass/pubs/catalog2004.pdf
Research:
http://www.nass.usda.gov/research/SERS.htm
Search words: national agricultural statistical service;
nass
Speaker here: Carol House
MSTS/2004
ANATOMY # 81
Slide 82
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Park Inventory and Monitoring Program
Program site:
http://science.nature.nps.gov/im/index.htm
Current Reports:
http://science.nature.nps.gov/im/reports.htm
Methods:
http://science.nature.nps.gov/im/standards.htm
Search words: national park monitoring natural
resources; nps im
Speaker here: Steve Fancy
MSTS/2004
ANATOMY # 82
Slide 83
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Resources Inventory( NRI)
Program site:
http://www.nrcs.usda.gov/technical/NRI/
Current Report:
http://www.nrcs.usda.gov/technical/land/nri02/
Methods:
http://www.statlab.iastate.edu/survey/nri/#Nussera
ndGoebel
Search words: national resources inventory;
web path: nrcs to technical to NRI
Speakers here: Wayne Fuller & Jeff Goebel
MSTS/2004
ANATOMY # 83
Slide 84
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Wetlands Inventory (NWI)
Program site:
http://wetlands.fws.gov/
Current Report:
http://training.fws.gov/library/Pubs9/wetlands86
-97_highres.pdf
Methods:
documented in above report
Search words:
national wetlands inventory
Speaker here: Tom Dahl
MSTS/2004
ANATOMY # 84
Slide 85
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
LEARNING MATERIALS RELATED TO
ENVIRONMENTAL SAMPLING
(From a course at Oregon State University)
ST571
http://oregonstate.edu/instruct/stat/urquhart/st5
71/index.htm
MSTS/2004
ANATOMY # 85
ANATOMY OF SAMPLING STUDIES
OF
ECOLOGICAL RESPONSES THROUGH TIME
{on the web at http://www.oregonstate.edu/instruct/st571/urquhart/anatomy/index.htm}
by
N. Scott Urquhart
Oregon State University, USA
and
Anthony R. Olsen
US EPA
MSTS/2004
ANATOMY # 1
Slide 2
THE AUTHORS
N. SCOTT URQUHART
Trained in Statistics
About 40 Years of Experience in Applications
Worked With Ecologists in Desert, Arctic, Pacific Northwest
Many Surveys with Rural Sociologists and Ag Economists
Including 10 years with EPA’s Environmental Monitoring and
Assessment Program (EMAP)
ACADEMIC And AGENCY; PLANT And ANIMAL
ANTHONY (Tony) R. OLSEN
Trained in Statistics
30+ Years of Experience in Private and Government Applications
Worked With Atmospheric Modelers And Air Pollution Field
Scientists
Survey Experience With Health Professionals And Large-scale
National Resource Monitoring
Now Statistical Lead with EPA’s EMAP
MSTS/2004
ANATOMY # 2
Slide 3
EVOLUTION OF THE “ANATOMY”
The first step in the development of the
ANATOMY focused on experimental design
situations.
Served as the structure for several part-semester
courses in advanced statistical methods at New
Mexico State University
Eventually published as
Urquhart, N. S. (1981). Anatomy of a study.
HortScience 16:621-627.
Experience with EMAP led to its expansion to
surveys
MSTS/2004
ANATOMY # 3
Slide 4
TODAY’S CONTEXT for
SURVEYS
“EMAP-type Situations”
EMAP = US EPA’S Environmental
Monitoring and Assessment Program
Estimate Status, Changes ...
In Indicators
Estimate Status, Changes, ...
In Extent
Describe Associations ...
MSTS/2004
ANATOMY # 4
Slide 5
Objective #1: Estimate the status, changes and trends in
selected indicators of the condition of our Nation's
ecological resources on a regional scale
with known confidence
17.6%
± 10%
6.8%
± 6%
31.8%
± 8%
43.8%
± 12%
Hypereutrophic
(N=258)
Eutrophic
Mesotrophic
Oligotrophic
Source: EMAP Northeast
MSTS/2004
Lakes Study
ANATOMY # 5
Slide 6
Objective #2: Estimate the status, changes and trends
in the extent and geographic coverage of our Nation's
ecological resources on a regional scale with known
confidence
12000
Est.
Lake #
North
east
SE
Est.
Area
SE
10000
8000
11,455
1,251
4,030
814
6000
Adir
1,506
285
1,082
395
4000
NEU
C/L/P
5,689
4,280
1,206
1,048
2,099
758
850
254
2000
0
Northeast
Adir = Adirondacks;
NEU = New England Uplands;
MSTS/2004
C/L/P
= Coast & Lake Plains
Adir
NEU
C/L/P
Source: EMAP NortheastANATOMY
Lakes Study
#6
Slide 7
Objective #3: Describe associations
between indicators of anthropogenic stress
and indicators of condition
Fish Index of Biotic Integrity
Relative Ranking of Stressors
Good
Sedimentation
(Insufficient
Data)
17%
Riparian Habitat
24%
Mine Drainage
17%
31%
11%
10%
Tissue Contamination
Fair
Phosphorus
5%
Nitrogen
5%
Acid Mine Drainage
0%
Proportion of Stream Length
Source: EMAP Mid-Atlantic
MSTS/2004
14%
Acidic Deposition
36%
Poor
25%
Highlands Assessment
1%
10%
20%
30%
40%
% of Stream Length
ANATOMY # 7
Slide 8
WHO MUST COMMUNICATE
MSTS/2004
Ecologists & Other Biologists
Statisticians
Geographers
Geographic Information Specialists (GIS)
Information Managers
Quality Assurance Personnel
Managers, At Various Levels
ANATOMY # 8
Slide 9
“SAMPLING”
A WORD OF MANY MEANINGS
Statisticians Often Associate It With Survey
Sampling
An Ecologist May Associate It with the Selection of
Local Sites or Material
A Laboratory Scientist May Associate It With the
Selection Of Material to be Analyzed from
Material Supplied
Common General Meaning, Varied Specific
Meanings
MSTS/2004
ANATOMY # 9
Slide 10
THE SPECIAL NEED
Communication Demands a Distinction
Between
The
Local Process of Evaluating a Response,
and
The Statistical Selection of a Sampling Unit,
For example,
A LAKE
A POINT ON A STEAM
MSTS/2004
ANATOMY # 10
Slide 11
THE SPECIAL NEED - continued
The Terms
Response Design
Sampling Design or Survey Design
Can Be Used to Make this Distinction
But a Complex Ecological Survey
Clearly Has More Parts Than These!
MSTS/2004
ANATOMY # 11
Slide 12
BASIC ROLES
Survey Design Tells Us Where to Go to
Collect Sample Information or Material
Response Design Tells Us What to Do
Once We Get There
But These Two Components Exist in a
Broader Context
MSTS/2004
ANATOMY # 12
Slide 13
AN IMPORTANT DISTINCTION
Monitoring Strategy
Conceptual
Impacted by Objectives
Addressable Without Regard to
the Inference Strategy
Inference Strategy
MSTS/2004
ANATOMY # 13
Slide 14
AN IMPORTANT DISTINCTION
- continued
Monitoring Strategy
..........
Inference Strategy
Places to Evaluate the Response – “the
WHERE”
Relation Between Points Evaluated and
the Population
IE, the Basis for Inference
MSTS/2004
ANATOMY # 14
Slide 15
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
Monitoring Strategy
Universe Model
Statistical Population
Domain Design
Response Design
These components
exist regardless
of the
inference strategy
Inference Strategy
Survey Design
Temporal Design
These components
exist for any
monitoring strategy
Quality Assurance Design
MSTS/2004
ANATOMY # 15
Slide 16
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 16
Slide 17
The MONITORING STRATEGY
The MONITORING STRATEGY MUST
RESPOND TO
Monitoring Objectives
State of Knowledge in Ecological Sciences
Characteristics of Ecological Resource(s)
of Interest
EXPECTED FUNDING Compared To COSTS
Operational Constraints
MSTS/2004
ANATOMY # 17
Slide 18
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 18
Slide 19
The UNIVERSE MODEL
Reality (Universe): Ecological Entity Within
a Defined Geographic Area to Be Monitored
Model of the Universe:
Development of a Monitoring Approach
Requires Construction of a Model for the
Universe
Elements Of The Universe Model: Set of
Entities Composing The Entire Universe
MSTS/2004
ANATOMY # 19
Slide 20
The UNIVERSE MODEL
Population Description and Its Sampling
Require Definition of the “Units” in the
Population
Discrete Units:
Lakes May Be Viewed This Way
Continuous Structure in Space of Some
Dimension:
2-space: forests or agroecosystems
1-space: Streams
3-space: Ground Water
MSTS/2004
ANATOMY # 20
Slide 21
THE MODEL FOR STREAMS
Strahler Orders
Second
Order
First Orders
First Orders
First Orders
First Order
MSTS/2004
ANATOMY # 21
Slide 22
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 22
Slide 23
The STATISTICAL POPULATION
The Collection of Units (as modeled) Over Some
Region of Definition
Spatial
Temporal
SPATIAL And TEMPORAL
Population Definition Could Include Features
Which Depend on Response Values
EX: acid sensitive streams at upper elevations
MSTS/2004
ANATOMY # 23
Slide 24
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 24
Slide 25
The DOMAIN Design
Specifies Subpopulations or “Domains” of
Special Interest
May Specify Meaningful Comparisons
Between Domains
Similar to Planned Comparisons in
Experimental Design Situations
Domain Design May Depend on Response
Values
Ex: Warm Versus Cold Water Lakes
MSTS/2004
ANATOMY # 25
Slide 26
The DOMAIN DESIGN - continued
Specifies Subpopulations or
“Domains” of
Special Interest
Determined From Defining Factors For
The Monitoring Activity
Must Have Critical Connection To Clients
Other Domains May Be Used For Analysis,
Without Having Been Used In Defining The
Monitoring Strategy
EX: EMAP domains include ECOAREAS
and STANDARD FEDERAL REGIONS
MSTS/2004
ANATOMY # 26
Slide 27
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 27
Slide 28
The RESPONSE Design
The Response Design Specifies
The Process Of Obtaining A Response
At An Individual Element (Site)
Of The Resource
During A Single Monitoring Period
Response: What Will Be Determined
On An Element –
Needs To Be Responsive to the Objectives of the
Monitoring Activity
MSTS/2004
ANATOMY # 28
Slide 29
The RESPONSE Design - continued
EMAP Responses Focus On Indicators of
STRESS and
Condition
The Response Design Also Defines
Plot Design
Measurement Protocols
Support Region – area around the site where material
is collected, or measurements are taken
Data Reduction Protocols
Calculation Of The Final Indicator Value for the
Element
MSTS/2004
ANATOMY # 29
Slide 30
The RESPONSE Design
- Continued
For example, consider a response related to
macroinvertebrates in streams
MSTS/2004
RESPONSE = proportion EPT (This is the
proportion of collected macrobenthos
organisms, mainly insects, which fall in the
taxonomic classes of Ephemeroptera ,
Plecoptera , or Trichoptera. Low values
indicate polluted streams; high values indicate
rather pristine streams)
ANATOMY # 30
Slide 31
The RESPONSE Design - continued - 2
... response related to macrobenthos ...
The COLLECTION UNITS could be 10
30cm x 30cm areas, systematically organized, at
the stream site, sampled with a “Surber
sampler”
The EVALUATION UNIT could be a jar
containing the composite of all macroinvertebrate
organisms collected at the 10 collection sites, or
MSTS/2004
The EVALUATION UNIT also could be a jar
containing a 1/6 subsample of the composite of
macroinvertebrate organisms collected in the
10 collection units.
ANATOMY # 31
Slide 32
The RESPONSE Design - continued - 3
... response related to macrobenthos ...
The LABORATORY EVALUATION of the
material would consist of determining and
recording the taxa (like family, genus, or
species) of each organism in the evaluation
material
The RESPONSE would be determined by
computing the number of organisms in the
evaluation material belonging to the E, P, T
taxonomic classes, and dividing this by the
number of organisms classified.
ANATOMY # 32
MSTS/2004
Slide 33
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 33
Slide 34
The INFERENCE STRATEGY
Is The Basis For Scientific Inference
Provides The Connection Between Objectives and the
Monitoring Strategy
Monitoring Strategy Usually Must Rely on Obtaining
Information on a Subset Of All Possible Elements in
the Universe
Specifies Which Elements of the Universe Will Have
Responses Determined on Them
Can Be Based On Either ... (continued )
MSTS/2004
ANATOMY # 34
Slide 35
The INFERENCE STRATEGY
(continued)
... Connection ...
... Subset ...
... Have Responses
Can Be Based On Either
Judgment Selection Of Units
Inferential Validity Rests on Knowledge Of Relation
Between the Universe And the Units Evaluated
Probability Selection Of Units
(The Focus Here)
MSTS/2004
ANATOMY # 35
Slide 36
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 36
Slide 37
The SURVEY Design
Probability Based Survey Designs Are
Considered Here
May Be Somewhat Limited To Sedentary Resources
Positive Features
(As An Observational Study)
MSTS/2004
Permit Clear Statistical Inference to
Well-Defined Populations
Measurements Often can be Made in Natural
Settings, Giving Rise to Greater Realism
Eventual Results
ANATOMY # 37
Slide 38
The SURVEY DESIGN - CONTINUED
Disadvantages
Limited Control Over Values of Predictor
Variables
Restricts Causative Inference
Usually Will Produce Inaccessible Sampling
Points
Good - For Inference
Bad - For Logistics
MSTS/2004
ANATOMY # 38
Slide 39
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
MSTS/2004
ANATOMY # 39
Slide 40
VIEW DOWN TRANSECT AT
MILE 12.3
MSTS/2004
ANATOMY # 40
Slide 41
CLIFF AT
MILE 135.2
(PARTIAL
HEIGHT)
MSTS/2004
ANATOMY # 41
Slide 42
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
Example: Frame for selecting field sites on streams
in the Western US
MSTS/2004
ANATOMY # 42
Slide 43
FRAME ERRORS
TO BE DOCUMENTED SHORTLY
Water Body Size
Flow Status -- re Perennial
Identified As Perennial, but not correct
Wastes Effort Of Field Crews
Identified as Non-perennial, but Really is
Perennial
Missed Resource
Inaccurate Assessment
MSTS/2004
ANATOMY # 43
Slide 44
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 44
Slide 45
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 45
Slide 46
EMAP-West Stream/river Length
(km ± 95% CI)
from Peck, et al (2002) - EMAP symposium
Fram e S ource
R F 3 F ram e
S iz e
RF3 Coded
P e re n n ia l
6 5 6 ,7 0 6
(P e re n n ia l S u rve y )
RF3 Coded
N o n - p e re n n ia l
1 ,6 2 8 ,9 8 0
(N o n - p e re n n ia l S u rve y )
T o ta l
MSTS/2004
2 ,2 8 5 ,6 8 6
E va lu a te d
E va lu a te d
“P e re n n ia l” N o n - p e re n n ia l
5 0 1 ,0 6 0
1 2 8 ,3 2 8
1 5 ,5 9 0
1 2 ,7 0 9
1 1 2 ,5 3 7
1 ,4 6 9 ,2 7 7
2 1 ,2 7 8
6 3 ,5 1 5
6 1 3 ,5 9 7
1 ,5 9 7 ,6 0 5
2 6 ,3 7 8
6 4 ,7 7 4
ANATOMY # 46
Slide 47
The SURVEY DESIGN - CONTINUED
Execution of a Sampling Plan Requires
A Sampling Frame
A way to identify elements in the population
Usually somewhat inaccurate for ecological
resources
– Example selecting vegetation sites along the Colorado
River in the Grand Canyon
Example: Frame for selecting field sites on streams
in the Western US
May change over time
– As, for example, land use changes
MSTS/2004
ANATOMY # 47
Slide 48
SITE SELECTION
Needs to Accommodate Realities Such As
Frame Imperfection
Frame Which Changes Over Time
Sites nearly Uniform Over the Resource
But with substantial randomization
Supports Variable Probability of Selection
Generalized Randomized Tessellation Stratified
Sampling = GRTS
The topic of the next session
MSTS/2004
ANATOMY # 48
Slide 49
The EMAP SURVEY Design
Assures Representation and Inference to
Populations
Adapted to Resource Characteristics
Emphasizes Spatial Allocation of Samples
Uses Two-phase Sampling; Phase I
Based on a Randomized Point Grid and
Associated Areas
MSTS/2004
ANATOMY # 49
Slide 50
EMAP EXAMPLE OF SELECTED
SITES
Mid-Appalachian Highlands Stream Pilot
MSTS/2004
ANATOMY # 50
Slide 51
MSTS/2004
ANATOMY # 51
Slide 52
EMAP EXAMPLE OF SELECTED
SITES
Western Stream Pilot
Non-perennial/perennial survey
Survey of perennial streams
MSTS/2004
ANATOMY # 52
Slide 53
MSTS/2004
ANATOMY # 53
Slide 54
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 54
Slide 55
The TEMPORAL Design
The TEMPORAL DESIGN specifies the
pattern of revisits to sites selected by the
Survey Design
Sampled population units are partitioned into
one (degenerate case) or more PANELS.
Each population unit in the same panel has the
same temporal pattern of revisits.
Panel definition could be probabilistic or
systematic
Several temporal designs follow
MSTS/2004
ANATOMY # 55
Slide 56
TEMPORAL DESIGN:
ROTATING PANEL
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1
2
3
4
5
1
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
2
3
4
5
6
7
8
9
MSTS/2004
6
7
8
9
1 0 1 1 1 2 1 3 ...
X
X
ANATOMY # 56
Slide 57
TEMPORAL DESIGN:
ROTATING PANEL
A ROTATING PANEL DESIGN IS THE TEMPORAL
DESIGN USED BY THE NATIONAL AGRICULTURAL
STATISTICAL SERVICE (US - “NASS”)
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE
EXPERIMENTAL DESIGN SENSE
IT IS FAIRLY WELL SUITED FOR ESTIMATION
“STATUS,” BUT NOT NEARLY PARTICULARLY
POWERFUL FOR DETECTING TREND OVER
INTERMEDIATE TIME SPANS
MSTS/2004
ANATOMY # 57
Slide 58
TEMPORAL DESIGN:
SERIALLY ALTERNATING
T IM E PE R IO D ( ex: Y E A R S )
PA N E L
1
1
X
2
3
4
2
3
4
5
6
7
8
X
X
X
X
X
X
10 11 12 13 ...
X
X
X
9
X
X
X
THIS TEMPORAL DESIGN IS “UNCONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE. THIS
WAS
THE TEMPORAL DESIGN INITIALLY PRESCRIBED
FOR EMAP, BUT ...
MSTS/2004
ANATOMY # 58
Slide 59
TEMPORAL DESIGN:
AUGMENTED SERIALLY ALTERNATING
T IM E P E R IO D ( ex: Y E A R S )
PA N E L
1
2
3
4
5
6
7
8
9
0
X
X
X
X
X
X
X
X
X X
1
X
2
3
4
X
X
X
X
X
X
X
X
X
X
X
X
10 11 12 13 ...
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 59
Slide 60
TEMPORAL DESIGN:
PARTIALLY AUGMENTED
SERIALLY ALTERNATING
T IM E P E R IO D ( ex: Y E A R S )
PA N E L
1
2
3
4
5
0
X
X
X
X
X
1
X
2
3
4
6
7
8
X
X
10 11 12 13 ...
X
X
X
9
X
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED”
IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 60
Slide 61
TEMPORAL DESIGN:
SERIALLY ALTERNATING
WITH CONSECUTIVE YEAR REVISITS
T IM E PE R IO D ( ex: Y E A R S )
PA N E L
1
2
1
X
X
2
3
4
X
3
4
5
6
X
X
X
X
X
X
X X
7
8
9
10 11 12 13 ...
X
X
X
X
X
X
X
X
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE
EXPERIMENTAL DESIGN SENSE
BUT IT PROVIDES VISITS TO ONLY HALF AS MANY
SITES AS THE BASE SERIALLY ALTERNATING DESIGN
MSTS/2004
ANATOMY # 61
Slide 62
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH CONSECUTIVE YEAR
REVISITS
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1
1
X
1A
X
2
3
4
5
6
7
8
X
X
2
X
2A
X
X
9
X
X
X
X
X
3
X
3A
X
X
4
X
4A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
1 0 1 1 1 2 1 3 ...
X
X
X
X
X
X
THIS TEMPORAL DESIGN IS “CONNECTED” IN THE EXPERIMENTAL DESIGN SENSE
MSTS/2004
ANATOMY # 62
Slide 63
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
T IM E P E R IO D ( ex: Y E A R S )
PANEL
1
1
X
1A
X
X
X
X
X
X
X
X
X
X
X
X
X
X
MSTS/2004
2
3
4
5
6
7
8
X
9
10
11
12
X
X
13
...
X
2
X
X
2A
X
X
X
X
X
X
X
X
X
X
X
X
3
X
X
X
3A
X
X
X
X
X
X
X
X
X
X
X
X
4
X
X
X
4A
X
X
X
X
X
X
X
X
X
X
X
X
ANATOMY # 63
Slide 64
THE REVISIT SPLIT PANEL OF THE
THE TEMPORAL DESIGN IN USE BY
EMAP - SURFACE WATERS
T IM E P E R IO D ( ex : Y E A R S )
PA N EL
1A
2A
3A
4A
MSTS/2004
1
2
X
X
X
X
X
X
3
4
5
6
X
X
X
X
X
X
X
X
X
X
X
X
X
X
7
9
X
X
X
X
X
X
X
X
8
1 0 1 1 1 2 1 3 ...
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
ANATOMY # 64
Slide 65
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
THIS TEMPORAL DESIGN IS “CONNECTED” IN
THE EXPERIMENTAL DESIGN SENSE
IT ALSO SUPPORTS ESTIMATION OF THE
SITE by TIME PERIOD (SITE by YEAR)
INTERACTION
REVISITS TO ABOUT 10% OF SITES ALLOCATES
ABOUT 30% OF RESOURCES TO REVISITS
MSTS/2004
ANATOMY # 65
Slide 66
TEMPORAL DESIGN: SPLIT PANEL
SERIALLY ALTERNATING
PLUS SERIALLY ALTERNATING WITH WITHIN YEAR
AND CONSECUTIVE YEAR REVISITS
THE TEMPORAL DESIGN IN USE BY EMAP - SURFACE WATERS
THIS TEMPORAL DESIGN IS “CONNECTED” IN
THE EXPERIMENTAL DESIGN SENSE
IT ALSO SUPPORTS ESTIMATION OF THE SITE by
TIME PERIOD (SITE by YEAR) INTERACTION
WE WILL RETURN TO THIS MATTER AFTER WE
CONSIDER THE QUALITY ASSURANCE DESIGN.
REVISITS TO ABOUT 10% OF SITES ALLOCATES
ABOUT 30% OF RESOURCES TO REVISITS
MSTS/2004
ANATOMY # 66
Slide 67
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
Universe Model
Statistical Population
Domain Design
Response Design
INFERENCE STRATEGY
Survey Design
Temporal Design
Quality Assurance Design
MSTS/2004
ANATOMY # 67
Slide 68
QUALITY ASSURANCE DESIGN
Defines Those Activities Intended to Provide Data
of Known Quality:
Blind duplicates
Accepted chemical standards, Etc
Can Provide Valid Estimates of the Variance of
Pure Measurement Error
MSTS/2004
ANATOMY # 68
Slide 69
RELEVANT COMPONENTS OF
VARIANCE FOR EMAP
POPULATION = LAKE or STREAM, for example
YEAR YEAR by SITE CREW SHORT TERM TEMPORAL - index window PROTOCOL ERROR -
MEASUREMENT ERROR -
MSTS/2004
ANATOMY # 69
Slide 70
RELEVANT COMPONENTS OF
VARIANCE FOR EMAP
WHAT DOES IT TAKE TO ESTIMATE THESE?
WHICH ONES SHOULD BE INCLUDED IN
“RESIDUAL” VARIANCE
WHICH TREND MUST OVERCOME TO BE
DEMONSTRATED?
MSTS/2004
ANATOMY # 70
Slide 71
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
UNIVERSE MODEL
(AS DISTINCT FROM THE UNIVERSE)
NOT RELEVANT
VIEWED AS A LIST
POPULATION
MSTS/2004
VERY SIMILAR, OTHER THAN DIFFERENCES
IMPLICIT IN THE UNIVERSE MODEL
ANATOMY # 71
Slide 72
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued
DOMAIN DESIGN
USUALLY NOT EXPLICITLY ACKNOWLEDGED,
BUT VIEWED AS PART OF THE DESIGN
PROCESS
RESPONSE DESIGN
QUESTIONNAIRE CONSTRUCTION AND DESIGN
SOME ELEMENTS OF INTERVIEW PROTOCOLS
SURVEY DESIGN
MSTS/2004
RELATIVELY SIMILAR
ANATOMY # 72
Slide 73
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued - 2
TEMPORAL DESIGN
PRESENT, BUT
MANY FINITE POPULATION SURVEYS ARE ONE-TIME
MANY LARGE SURVEYS HAVE A TEMPORAL
DIMENSION
CHANGE USUALLY IS OF FAR MORE INTEREST
THAN TREND
SURVEY LITERATURE DISTINGUISHES
BETWEEN
GROSS CHANGE - follows units across time
NET CHANGE - recognizes that change can occur
MSTS/2004
ANATOMY # 73
in
Slide 74
COMPARISONS TO USUAL PERSPECTIVES
OF FINITE POPULATION SAMPLING
continued - 3
QUALITY ASSURANCE DESIGN
SURVEYS OF HUMANS (OR BUSINESSES, ETC)
HAVE A VERY SIMILAR INTENT, BUT DETAILS
DIFFER SUBSTANTIALLY
SUPERVISION IN PHONE SURVEYS
QUESTIONS TO CONFIRM RESPONDENT
CONSISTENCY
REINTERVIEWS and SIMILAR
REEVALUATIONS
MSTS/2004
ANATOMY # 74
Slide 75
SAMPLING STUDIES OF ECOLOGICAL
RESPONSES THROUGH TIME HAVE
MONITORING STRATEGY
UNIVERSE MODEL
STATISTICAL POPULATION
DOMAIN DESIGN
RESPONSE DESIGN
These components
exist regardless
of the
inference strategy
INFERENCE STRATEGY
SURVEY DESIGN
TEMPORAL DESIGN
These components
exist for any
monitoring strategy
QUALITY ASSURANCE DESIGN
MSTS/2004
ANATOMY # 75
Slide 76
LINKS TO MONITORING
REPORTS/DOCUMENTATION
Links to various monitoring programs,
To the program’s site
One to its methods, and
One to a current report.
As links change in an unpredictable fashion,
The “search words” should provide a quick path to the
current link.
Speaker on this program representing that program
MSTS/2004
ANATOMY # 76
Slide 77
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Adelaide Coastal Waters Study
Program site:
http://www.environment.sa.gov.au/epa/acws.html
http://www.clw.csiro.au/acws/ (more technical of the two)
Current Report:
http://www.environment.sa.gov.au/epa/pdfs/acwsnewsno5.pdf
Methods:
http://www.clw.csiro.au/acws/IS1.html
Search words: adelaide coastal waters study; acws au
Speaker: David Fox, University of Adelaide
MSTS/2004
ANATOMY # 77
Slide 78
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Alberta Biodiversity Monitoring Program (ABMP)
Program site:
http://www.abmp.arc.ab.ca/
Current Report:
http://www.abmp.arc.ab.ca/AnnualReport2003.pd
f
Methods:
http://www.abmp.arc.ab.ca/ScienceProtocols.htm
Search words: alberta biodiversity monitoring
program; abmp ca
Speaker: None, unfortunately
MSTS/2004
ANATOMY # 78
Slide 79
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Environmental Monitoring and Assessment
Program (EMAP)
Program site:
http://www.epa.gov/emap/
Current Report:
http://www.epa.gov/maia/html/maha.html
Methods:
http://www.epa.gov/emap/html/pubs/docs/groupdocs/
surfwatr/field/ws_abs.html
Search words: epa environmental monitoring assessment
program; emap epa
Speaker here: Tony Olsen
MSTS/2004
ANATOMY # 79
Slide 80
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
Forest Inventory and Analysis (FIA)
Program site:
http://fia.fs.fed.us/
Current Report:
http://www.fs.fed.us/research/sustain/
(for example)
Methods:
http://fia.fs.fed.us/FIAProgramInformation.htm
(accessible at this link)
Search words: forest inventory analysis; FIA USDA
Speaker here: Mike Williams
MSTS/2004
ANATOMY # 80
Slide 81
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Agricultural Statistical Service
(NASS)
Program site:
http://www.usda.gov/nass/
Current Reports:
http://www.usda.gov/nass/pubs/catalog2004.pdf
Research:
http://www.nass.usda.gov/research/SERS.htm
Search words: national agricultural statistical service;
nass
Speaker here: Carol House
MSTS/2004
ANATOMY # 81
Slide 82
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Park Inventory and Monitoring Program
Program site:
http://science.nature.nps.gov/im/index.htm
Current Reports:
http://science.nature.nps.gov/im/reports.htm
Methods:
http://science.nature.nps.gov/im/standards.htm
Search words: national park monitoring natural
resources; nps im
Speaker here: Steve Fancy
MSTS/2004
ANATOMY # 82
Slide 83
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Resources Inventory( NRI)
Program site:
http://www.nrcs.usda.gov/technical/NRI/
Current Report:
http://www.nrcs.usda.gov/technical/land/nri02/
Methods:
http://www.statlab.iastate.edu/survey/nri/#Nussera
ndGoebel
Search words: national resources inventory;
web path: nrcs to technical to NRI
Speakers here: Wayne Fuller & Jeff Goebel
MSTS/2004
ANATOMY # 83
Slide 84
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
National Wetlands Inventory (NWI)
Program site:
http://wetlands.fws.gov/
Current Report:
http://training.fws.gov/library/Pubs9/wetlands86
-97_highres.pdf
Methods:
documented in above report
Search words:
national wetlands inventory
Speaker here: Tom Dahl
MSTS/2004
ANATOMY # 84
Slide 85
LINKS TO MONITORING
REPORTS/DOCUMENTATION
(Continued)
LEARNING MATERIALS RELATED TO
ENVIRONMENTAL SAMPLING
(From a course at Oregon State University)
ST571
http://oregonstate.edu/instruct/stat/urquhart/st5
71/index.htm
MSTS/2004
ANATOMY # 85