John Sweka USFWS Northeast Fishery Center Lamar PA What is the goal of fish habitat restoration efforts and partnerships To create morebetter fish habitat To enhancerecoverrestore fish populations ID: 720399
Download Presentation The PPT/PDF document "Designing Effective Monitoring Programs ..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Designing Effective Monitoring Programs for Fish Population Response to Habitat Restoration
John
Sweka
USFWS – Northeast Fishery Center
Lamar, PASlide2Slide3
What is the goal of fish habitat restoration efforts and partnerships?
To create more/better fish habitat
To enhance/recover/restore fish populationsSlide4
How do we know if we met our goal?
We need to monitorSlide5
Adaptive Management
DOI Technical Guide
Example Models:
More Large Woody Debris = More Brook Trout
Lower Sediment Load = More Brook TroutSlide6
Issues of spatial scale
Defining the population of interest
Depends on the goal of the habitat restoration work
Reach population vs. stream population vs. range wide population
Often a mismatch between the population of interest and spatial scale of habitat restoration and monitoring
Can lead to erroneous conclusions about the effectiveness of habitat restorationSlide7
300 m
Brook trout abundance
A
0
= 15
A
1
= 30
100% increase !!!!!!Slide8
300 m
6000 m
If abundance throughout the rest of the steam stays the same, this is only a 5% increase.Slide9
Issues of spatial scale
Life history information can inform decisions on spatial scale of monitoring
Home range – Does monitoring encompass the home range of individuals in the population?
Schooling behavior – Is the species of interest patchily distributed?
Migration – Does the species of interest migrate from one habitat to another while completing its life cycle?Slide10
A good example of monitoring at the population scale
Liermann
, M. and P.
Roni
. 2008. More sites or more years? Optimal study design for monitoring fish response to watershed restoration. North American Journal of Fisheries Management 28: 935-943.
“…the only way to assess the population
level effects
of watershed scale restoration is to monitor at the population level.”
Monitored salmon
smolt
migration from small streams with and without habitat restoration.
Employed knowledge of life history information (migration)
Replicated experiment with controlsSlide11
Issues of temporal scale
How does the duration of monitoring compare to the life history of the species of interest?
Generation Time – amount of time it takes one cohort to grow up and replace another; can be calculated from a life table or a Leslie matrix
YOY
Age1
Age2
Age3
Age4+
YOY
0.00
37.50
56.25
97.50
150.00
Age1
0.06
0.00
0.00
0.00
0.00
Age2
0.00
0.10
0.00
0.00
0.00
Age3
0.00
0.00
0.10
0.00
0.00
Age4+
0.00
0.00
0.00
0.10
0.01
Population growth rate(
λ
) = 1.6, generation time = 2.14Slide12
Issues of temporal scale
If habitat restoration has a population level effect, we would not expect to begin seeing any real change until the expected generation time is reached
Likely longer due to variation in other uncontrollable factors (e.g. rainfall, flow, temperature, predation etc.)
Length of monitoring can greatly influence conclusions that are drawnSlide13
Sweka
, J.A. and K.J. Hartman. 2006. Effects of large woody debris addition on stream habitat and brook trout populations in Appalachian streams. Hydrobiologia 559: 363-378.Slide14
Sweka
, J.A. and K.J. Hartman. 2006. Effects of large woody debris addition on stream habitat and brook trout populations in Appalachian streams. Hydrobiologia 559: 363-378.Slide15
Types of Monitoring Designs
Before-After Design
Time
Abundance
Assumes everything but the treatment remained constant through time
Best if there is many years of pre- and post- data
Simply compare pre- and post- mean abundanceSlide16
Types of Monitoring Designs
Pre/Post Pairs
Time
Abundance
Allows assessment of site-to-site variation
Temporal scale may not be long enough
Ignores any regional trends in abundance that may existSlide17
Types of Monitoring Designs
Before-After-Control-Impact Design (BACI)
Time
Abundance
Has an independent control site
Best if there are several years of pre- and post- data
Interested in the difference between treatment and control
Embraces natural variation
Control site
Treatment siteSlide18
Types of Monitoring Designs
Before-After-Control-Impact Design (BACI)
Time
Abundance
Control can have higher or lower abundance than the treatment site
Control site
Treatment siteSlide19
Types of Monitoring Designs
Before-After-Control-Impact Design (BACI)
Time
Abundance
Can have several treatment and control sites
Control sites
Treatment sitesSlide20
Types of Monitoring Designs
Real World Case
Time
Abundance
Control site
Treatment site
Limited funding, 1 year pre-, couple years post-
Add a control – separate natural variation from treatment variation
Continue monitoring treatment and control – look for treatment x time interactionSlide21
Power and Sample Size
Power
– the probability of correctly rejecting the null hypothesis of no change (no difference) when some specified alternative is correct
How much Power do you need?
Depends on the consequences
Law of diminishing returns – rate of increase in power decreases with increasing sample size
Increasing power can be costlySlide22
Power and Sample Size
How many samples should I take to detect a difference?
Choice of alpha (chance of falsely rejecting
H
o
)
Whether the test will be one- or two-tailed
Value of the alternative
H
a
(desired difference to detect)
Choice of design
Some assumption about the behavior of the variation in the data (
e.g
variance proportional to mean)
Estimate of the variation (standard
deveiation
or variance)
Slide23
Power and Sample Size
Guidelines for choosing an
effect size
(
Gerow
2007)
Small Effect
– the smallest difference that elicits your interest
Large Effect
– the smallest difference that you would definitely not want to fail to detect
Medium Effect
- the average of small and large effectsSlide24
Power and Sample Size
Gerow
, K. G. 2007. Power and sample size estimation techniques for fisheries management: Assessment and a new computational tool. North American Journal of Fisheries Management 27: 397 – 404.
Gerrodette
, T. 1987. A power analysis for detecting trends. Ecology 68: 1364-1372.Slide25
Conclusions
Effective monitoring starts with a clearly defined population of interest and goals
Incorporate the life history, home range, and behavior of the target species
Have a control and avoid
psuedoreplication
View monitoring as an experiment (hypothesis testing)
Use power analysis to inform study design
Funding timelines don’t match biological timelines
Additional partnerships
Creative ways to extend funding
Work with funding sources for changeSlide26
Questions??