Transcript Uzarski et al June 2010 IJC-WL update report to Val_final

Water Levels Impacting Great Lakes
Coastal Wetlands: An update of metric
development
Donald Uzarski, Mathew Cooper, and Brent Murry
June 14, 2010
Contents
• Invertebrate community metrics
– Les Cheneaux Islands Region of N. Lake
Huron
• Fish assemblage metrics
– Saginaw Bay
• Invertebrate and fish response to
vegetation zone loss
– Saginaw Bay and northern Lake Huron
Invertebrate community metrics
Background and Hypothesis
Many studies suggest that changes in hydrology do
produce significant changes in macrophyte community
composition. Results of our past published studies relate
macroinvertebrate community composition to dominate
vegetation types with pronounced differences among types.
Here we attempt to develop indicators while keeping
vegetation type and depth relatively constant, therefore
isolating the effect of annual water level. We hypothesized
a shift in macroinvertebrate community composition related
to water level change and direction that was independent of
depth and dominant vegetation type. A shift of this nature
could be used to infer fine scale changes in ecosystem
structure and function related to hydrology.
Methods - Invertebrates
• Study Sites
– Les Cheneaux Islands Region of N. Lake Huron.
– 10 Fringing Coastal Wetlands.
• Macroinvertebrate Samples
–
–
–
–
–
Collected from Schoenoplectus 1997 – 2002
Water Levels Declined ~ 1 m.
Followed Migration of Plant Zone w/Declining Water
D-Frame Dip Net
3 Replicates Per Site Per Year
• Data Analysis
– NMDS of Most Data Rich Site (Mackinac Bay)
– Pearson Correlation – Mean Water Level and Dim 1
– Determined Abundant Taxa Most Responsible
Mackinaw Bay the most data-rich site
Invertebrate Community Compostions and 1997-2002 Water Levels
Mackinac Bay, Northern Lake Huron
Scrapers
Invertebrate Community Composition
(NMDS Dimension 1 (48%))
1.5
1.0
Significant Correlation
between NMDS Dim 1 and
Water Levels
Pearson Correlation
r = -0.823; p = 0.044
0.5
Keeping Plant Zone
Constant
Keeping Depth
Constant
0.0
-0.5
Caenidae and Asellidae
Weighted Heaviest in the
Relationship
-1.0
-1.5
175.8
Shredders
Collector/Gatherers
176.0
176.2
176.4
176.6
176.8
Lake Huron/Michigan Water Levels (m)
177.0
177.2
Invertebrate Community 1997-2002 Water Levels
Mackinac Bay, Northern Lake Huron
70
Pearson Correlation
r = 0.823
p = 0.044
60
Caenidae Numbers
50
40
30
20
10
0
175.8
176.0
176.2
176.4
176.6
176.8
177.0
177.2
Lake Huron/Michigan Water Levels (m)
The collector/gatherer Caenidae Populations Seems to Reflect the Previous
Years Water Levels
Invertebrate Community 1997-2002 Water Levels
Mackinac Bay, Northern Lake Huron
70
Pearson Correlation
r = 0.683
p = 0.135
60
Asellidae Numbers
50
40
30
20
10
0
175.8
176.0
176.2
176.4
176.6
176.8
177.0
177.2
Lake Huron/Michigan Water Levels (m)
The Shredder Asellidae Populations Seems to Reflect the Previous
Years Water Levels
Lakes Michigan and Huron
At all 10 Les Cheneaux fringing wetland sites
177.2
177.0
Water Levels (m)
176.8
176.6
Invert Data Seem to Reflect
the Previous Year
176.4
176.2
176.0
175.8
1995
1996
1997
1998
1999
Year
2000
2001
2002
2003
Invertebrate Community 1997-2002 Water Levels
Northern Lake Huron (10 Sites)
25
Pearson Correlation
r = 0.801
p = 0.055
Caenidae Mean Numbers
10 Sites
20
15
10
5
0
175.8
176.0
176.2
176.4
176.6
176.8
177.0
177.2
Lake Huron/Michigan Water Levels (m)
The collector/gatherer Caenidae Populations Seems to Reflect the Previous
Years Water Levels
Invertebrate Community 1997-2002 Water Levels
10 Northern Lake Huron Sites
40
Pearson Correlation
NS
Asellidae Mean Numbers
35
30
25
20
15
10
5
175.8
176.0
176.2
176.4
176.6
176.8
177.0
177.2
Lake Huron/Michigan Water Levels (m)
The Shredder Asellidae Populations Seems to Reflect the Previous
Years Water Levels
Discussion – invertebrate metrics
Water depth and the dominant vegetation type was kept constant yet
there was still a significant relationship between invertebrate
community composition and water levels. This was likely driven, in
part, by a shift from a detritus based food web to and algal based food
web. As water levels rose, more protected areas with denser
vegetation were inundated with water. Through time, the deeper water
with more hydrologic energy likely reduced vegetation density resulting
in an abundance of detritus favoring shredders and collector/gatherers.
As water levels declined, areas of sparse vegetation became benign
allowing sunlight to penetrate gradually favoring algae. During
declining water level years, scrapers trended towards becoming more
abundant, but there was no significant relationship between their
numbers and water levels.
Conclusions – invertebrate metrics
• Macroinvertebrates metrics that respond to changes in water level
and direction independent of large scale vegetation shifts can be
developed.
• There appears to be a shift from shredders and collector/gatherers
to grazers as water levels decline.
• This shift may be the result of the wetland moving from a detritus
based food web in response to rising water levels to an algal based
food web as water levels decline.
• Rising and falling water levels result in a temporally diverse
macroinvertebrate community as well as dynamic ecosystem
structure and function.
Fish assemblage metrics
• Water level drivers derived from several
components of the natural flow regime concept
• Fish assemblage attributes investigated thus far
include:
– Total mean abundance
– Species richness
– Simpson evenness
NATURAL FLOW REGIME
MAGNITUDE
WATER
QUALITY
FREQUENCY
DURATION
ENERGY
SOURCES
TIMING
PHYSICAL
HABITAT
RATE OF CHANGE
BIOTA
ECOLOGICAL INTEGRITY
(modified from Poff et al. 1997)
Water Level Metrics
• Magnitude * Timing
– Logic
• Water elevation (magnitude) affects wetland area which will
influence the species abundance and composition
• Monthly changes (timing) in WL elevation will differentially
affect habitat use of species depending on their unique life
histories
– Metrics:
• Monthly (March – Aug.) mean, min., max. water level (m)
• Monthly (March – Aug.) change (STDEV)
Water Level Metrics
• Rate of Change
– Logic:
• Environmental cues including (primarily) photoperiod,
temperature, and (secondarily) water level changes initiate
spawning activity in fish and potentially metamorphosis and
emergence of some invertebrates
• Broad seasonal changes and timing relative to photoperiod
(equinox and solstice key photoperiod endpts.) are potentially
important cues
– Metrics
• Winter to summer rate of WL change
– (July mean WL - Jan Mean WL) / # days
– assumed Jan 15th and July 15th for # days count, leap years = 182
(2000, 2004, 2008), otherwise 181
• Rate of WL change spring equinox to summer solstice
– Calc as June mean WL - March mean WL / 92 days (March 21
to June 21 = 92 days)
Saginaw Bay Fish Data – Primary Response
Variables
• Total mean CPUE
– GLCWC standardized trap-nets
– Inner and outer Schoenoplectus zones
• Species Richness
– Not presently rarified, very different # fish among sites
but same effort
– Inner and outer Schoenoplectus zones
• Simpson Evenness
– Inner and outer Schoenoplectus zones
• Simpson Diversity
– Inner and outer Schoenoplectus zones
Sample distribution among years and zones
Ecoregion region
Habitat Zone
Year
Total #
nets
#Nets for Model
Development
# Nets for model
validation
Saginaw Bay
Inner
Schoenoplectus
2002
12
8
4
Saginaw Bay
Inner
Schoenoplectus
2003
3
3
0
Saginaw Bay
Inner
Schoenoplectus
2004
9
6
3
Saginaw Bay
Inner
Schoenoplectus
2006
12
8
4
Saginaw Bay
Inner
Schoenoplectus
2008
6
4
2
Saginaw Bay
Outer
Schoenoplectus
2002
12
8
4
Saginaw Bay
Outer
Schoenoplectus
2003
12
8
4
Saginaw Bay
Outer
Schoenoplectus
2004
7
5
2
Saginaw Bay
Outer
Schoenoplectus
2006
12
8
4
Saginaw Bay
Outer
Schoenoplectus
2008
9
6
3
Annual sites sampled and effort
Ecoregion
Year
Site Name
Total # nets
Saginaw Bay
2002
Bradleyville
3
Saginaw Bay
2002
Pinconning
6
Saginaw Bay
2002
Vanderbilt Park
6
Saginaw Bay
2002
Wigwam Bay
6
Saginaw Bay
2002
Wildfowl Bay
3
Saginaw Bay
2003
Almeda Beach
3
Saginaw Bay
2003
Nyanquing
3
Saginaw Bay
2003
Vanderbilt Park
3
Saginaw Bay
2003
Wigwam Bay
6
Saginaw Bay
2004
Wigwam Bay
3
Saginaw Bay
2004
Bayport
4
Saginaw Bay
2004
Linwood Beach
3
Saginaw Bay
2004
Nyanquing
3
Saginaw Bay
2004
Whites Beach
3
Saginaw Bay
2006
Pinconning
6
Saginaw Bay
2006
Sebawing
6
Saginaw Bay
2006
Wigwam Bay
6
Saginaw Bay
2006
Vanderbilt Park
6
Saginaw Bay
2008
Pinconning
6
Saginaw Bay
2008
Vanderbilt Park
3
Saginaw Bay
2008
Wigwam Bay
6
Outer Schoenplectus
Total Mean Fish Abundance
Mean Total Fish Abundance = -70,897 + 398.53526 * MaxMayWL – 20,629 *
STDEVMayWL + 46,926 * STDEVJulyWL
Model based on 5 years: 2002, 2003, 2004, 2006, 2008
Adj. r2 = 0.99, F3,4 = 13,036.2, P = 0.0064
•
•
Predictions tend to underestimate abundance
Strong correlation between observed and predicted, but there is 1 high
leverage point (influential)
Predicted mean total fish
abundance (CPUE)
•
•
900
800
1:1 line
700
600
500
400
300
200
R2 = 0.928
100
0
0
200
400
600
800
Observed mean total fish abundance (CPUE)
1000
Observed species richness
outer Schonoplectus
Outer Schoenplectus Fish Species
Richness
Model based on five years: 2002,
2003, 2004, 2006, 2008
20
18
16
SpR = 32.02170 – 4,918.77561 *
Rate of WL increase
Spring Equinox – Summer
Solstice
14
12
R2 = 0.90, F1,4 = 26.60
P = 0.0141
10
8
6
Predicted species richness
outer Schoenplectus zone
0
0.001
0.002
0.003
0.004
0.005
Rate of WL increase spring equinox to summer solstice
(m/day)
• Moderate correlation between
observed and predicted species
richness
• model tended to over-estimate
species richness
25
20
1:1 line
R2 = 0.59
15
10
5
0
0
5
10
15
20
Observed species richness outer Schoenoplectus zone
25
Outer Schoenplectus
Fish Assemblage Simpson Evenness
Evaluation based on 5 years: 2002, 2003, 2004, 2006, 2008
Simpson Evenness = no suitable model found
– The only significant model showed high multicollinearity and poor
observed/predicted concordance
Summary of water level influence on fish
abundance and diversity – Outer Schoenplectus
Fish response
WL attribute
Direction of
influence
Total mean
abundance
May max. WL
May WL STDEV
July WL STDEV
+
+
Species richness
Rate of change
Equinox – solstice
-
Simpson evenness
n/a
n/a
Inner Schoenplectus
Total Mean Fish Abundance
Mean Total Fish Abundance = 422.43192 + 1,726.87296 * AprilWLSTDEV –
5,366.62844 *MayWLSTDEV – 249.52950 * JulyWLRange
Model based on 5 years: 2002, 2003, 2004, 2006, 2008
Adj. r2 = 1.00; F3,4 = 5,688,984; P = 0.0003
•
Three of four years predictions were near 1:1 line, but a fourth year (2004)
was predicted far lower than observed resulting in a relatively weak
correlation between observed and predicted
Predicted mean total fish
abundance (CPUE)
•
•
250
200
1:1 line
150
100
2
R = 0.39
50
0
0
50
100
150
200
Observed mean total fish abundance (CPUE)
250
Inner Schoenplectus
Fish Assemblage Species Richness
Evaluation based on 5 years: 2002, 2003, 2004, 2006, 2008
Species richness = no suitable model found
– A significant model was found but predictions were extremely poorly correlated to
observed values (r2 = 0.01)
Inner Schoenplectus
Fish Assemblage Simpson Evenness
Fish Simpson Evenness = 8.21008 – 0.04826*JuneMinWL + 15.70754 *
MayWLSTDEV – 2.05742 * AugWLrange
•
•
Model based on 5 years: 2002, 2003, 2004, 2006, 2008
Adj. r2 = 1.00; F3,4 = 3.69x109; P < 0.0001
•
Strong correlation between observed and predicted, though one data point
(2008) again appears to have a strong influence on this relationship
Points are well distributed along 1:1 line
0.60
Predicted Simpson
evenness
•
0.50
R2 = 0.75
0.40
0.30
0.20
0.10
0.00
0
0.1
0.2
0.3
0.4
Observed Simpson evenness
0.5
0.6
Summary of water level influence on fish
abundance and diversity – Inner Schoenplectus
Fish response
WL attribute
Direction of
influence
Total mean
abundance
April WL STDEV
May WL STDEV
July WL range
+
-
Species richness
n/a
n/a
Simpson evenness
June min. WL
May WL STDEV
Aug WL range
+
-
Conclusions
• Variance in monthly water levels, in particular May, appear to have the most
impact on fish abundance and diversity.
– May WL variance was negatively associated with fish abundance in both the inner and
outer Schoenoplectus zones, but was positively associated with fish assemblage
evenness in the inner zone.
– Water level variance likely influences the habitat quality perceived by spawning fish
during and may limit reproductive effort and/or success (short-term stranding of eggs in
shallow water during low water events, seiches combined with low levels).
– Greater WL variance in July and August is associated with lower total abundance and
evenness in the inner zone, but higher abundance in the outer zone.
• This likely reflects fluctuations in the inner zone that reduces water depth and habitat quality
pushing fishes into deeper water. Fewer species remain in the inner zone during periods of
high fluctuations reducing evenness (i.e. dominance of a few species that remain under
generally unfavorable conditions).
• Species richness in the outer zone was negatively related to the rate of WL
increase between the spring equinox and the summer solstice.
– The rate of WL change relative to photoperiod (i.e. equinox and solstice) and water
temperature is a common cue to many fish species to initiate spawning and other life
history actions.
– As the rate of WL change increases this may alter habitat suitability conditions, reduce
spawning activity or cause young-of-the-year fish to move out of the wetland areas
during nursery, alternatively it may allow greater predation pressure as higher water
levels might increase large fish use of this habitat type for foraging.
Invertebrate and fish community response to vegetation
loss or zone contraction
A likely response to a reduction in water level variability would be the loss of
emergent vegetation or a contraction of emergent vegetation zones. Such
changes reduce the area of inundated vegetated habitat for fish and
invertebrates. Since macrophytes provide a number of critical resources for
fauna, we predict a reduction in the productivity and diversity of these groups
if water level variability is reduced and vegetation zones contract. Results
from one published study (Uzarski et al. 2009) and one unpublished study
(Cooper et al. In Prep) comparing faunal community composition from
vegetated and adjacent unvegetated areas provide insight on the likely
consequences of such changes.
Methods:
• Seven paired (vegetated and unvegetated) sites in Saginaw Bay. Vegetation was
dominated by bulrushes. Sampling was conducted in Summer, 2005.
• Triplicate timed dip-net samples. Nets were swept through upper sediment layer for
three minutes. Taxa were sorted to lowest operational taxonomic unit in the
laboratory.
Results:
• Unvegetated habitats dominated by small taxa, especially midge larvae
(Chironomidae), biting midge larvae (Bezzia), and water mites (Hydracarina)
while vegetated sites were dominated by larger taxa such as amphipods and
a number of snail taxa.
• Taxon richness in the vegetated habitats (17.9 ± 1.0) was significantly higher
(p = 0.031) than in the open water zones (9.5 ± 1.1).
Conclusions:
• A loss or contraction of vegetated habitat is very likely to have significant impacts
on macroinvertebrate community structure. These changes are likely to also
affect organisms higher in the food chain such as fish and birds.
from Cooper et al. (In Prep):
GREAT LAKES COASTAL MARSH FRAGMENTATION: EDGE AND AREA
EFFECTS ON ZOOPLANKTON, MACROINVERTEBRATE, AND
LARVAL FISH COMMUNITIES
Comparison of zooplankton, larval fish, and macroinvertebrate communities
between vegetated and adjacent unvegetated habitats
Methods:
•Quatrefoil light traps were set overnight in 16 fringing marshes of Lake Huron
• Compared vegetated and unvegetated habitats.
• Sampling conducted in Summer of 2005.
Results:
Mean ( ±SE):
Unvegetated
Vegetated
Paired t-tests
p:
Zooplankton
CPUE
17,356±6,684
87,775±47,836
0.19
6.6±0.9
6.0±0.8
0.07
1.17±0.10
1.08±0.20
0.64
0.168±0.067
0.552±0.265
0.17
Macroinvertebrat
es
CPUE
440±86
912±433
0.3
Richness
8.9±0.7
12.3±0.91
<0.01
0.88±0.10
1.27±0.13
0.01
7±2
47±20
0.08
1.3±0.3
2.7±0.4
<0.01
0.24±0.09
0.58±0.11
0.01
Richness
Shannon diversity
AFDM (g)
Shannon diversity
Larval fish
CPUE
Richness
Shannon diversity
Conclusions:
• Macroinvertebrate richness and Shannon diversity were significantly higher in
vegetated habitats relative to unvegetated habitats.
• Larval fish CPUE, richness and Shannon diversity were significantly higher in
vegetated habitat relative to unvegetated habitats.
• Changes to water level regime that result in a loss or constriction of vegetation
will likely have significant impacts on these faunal groups.