Upload
keiki
View
48
Download
0
Tags:
Embed Size (px)
DESCRIPTION
Ecological Thresholds for Salt Marsh Nekton and Vegetation Communities. Mary-Jane James-Pirri Graduate School of Oceanography University of Rhode Island Jeffrey Swanson Computer Science and Statistics University of Rhode Island Charles Roman National Park Service North Atlantic Coast - PowerPoint PPT Presentation
Citation preview
Ecological Thresholds for Salt Marsh Nekton and Vegetation Communities
Mary-Jane James-PirriGraduate School of OceanographyUniversity of Rhode Island
Jeffrey SwansonComputer Science and StatisticsUniversity of Rhode Island
Charles RomanNational Park ServiceNorth Atlantic Coast Cooperative Ecosystem Studies Unit
Howard GinsbergUSGS Patuxent Wildlife Research Center
James HeltsheComputer Science and StatisticsUniversity of Rhode Island
Nauset Marsh, Cape Cod NS
Project Background & Objectives• Funded by the USGS Status & Trends Program.• Use existing data to define ecological thresholds for Vital Signs associated with
the National Park Service’s Inventory and Monitoring Program.• Study Focus: Vital Signs of salt marsh nekton & vegetation community
composition from NPS Monitoring Protocols.• Identify patterns in community structure and correlate patterns with landscape
gradients (e.g., degree of urbanization), if possible.
The Marshes
Sites had a range of hydrologic disturbance:• No disturbance (reference sites)• Tidal restriction• Tidal restoration• Open marsh water management (OMWM) for mosquito control
~ 50% of sites were sampled in multiple years (mostly 2001 to
2006)
~ 50% of marshes were reference sites
Edwin B. Forsythe NWR
Prime Hook NWR
Assateague Island NS
Boston Harbor Islands NPACape Cod NS
Sagamore Hill NHS
Fire Island NSGateway NRA
Colonial NHS
George Washington Birthplace NM
Rachel Carson NWR
Parker River NWR
Stewart B. McKinney NWR
Long Island Complex NWR
Narragansett Bay, RI14 sites Dartmouth, MA sites
Sachuest Point NWR
National Parks
US Fish & Wildlife Refuges
Other sites
83 individual marshes in 48 watersheds
The DatasetsVegetation Community Data (70 sites, 156 individual datasets)
• Point intercept data from 1-m2 plots• 203 cover types – live vegetation & non-vegetative cover (water, wrack)
Nekton Community Data (81 sites, 185 individual datasets)• Relative abundance data from enclosure gear (throw traps & ditch nets)• 62 species - fish and crustaceans
Landscape metrics (48 sub-watersheds)• 2000 U.S. Census data (human population density of sub-watershed)• 2001 National Land Cover Data (land use within 5 km radius) (not presented)
Multivariate techniques were used to explore patterns in community composition
Cape Cod NS, MA Gateway NRA, NY Fire Island NS, NY
-6
-4
-2
0
2
4
6
-4 -2 0 2 4 6 8 10
Vegetation Community – Common cover types
• Reduced dataset to 21 common cover types (present in >20% of datasets)• 76 reference and 80 disturbed marsh datasets:
• Principal Component Analyses (PCA), 1st three principal components only accounted for ~40% (1st PC: 20%);
• No clear pattern in PCA scores using common cover types;• No clear gradient of land use in PCA scores;• No correlation with watershed human population density.
2nd P
rinci
pal C
ompo
nent
Reference sitesDisturbed sites
1st Principal Component
Gradient of Marsh Disturbance
Reference sites: reflect desired or target condition.
Disturbed sites: reflect degraded or less desired condition.
0
10
20
30
40
50
60
70
Perc
ent c
over
HydrologicallyAltered (OMWM)~ 11 species
Reference sites~13 species
Restoring sites~35 species
Restricted sites~26 species
Vegetation Groups- Wetland status & salinity tolerance
• Grouped vegetation by wetland status & salinity tolerance using USDA definitions (e.g., obligate wetland species, high salinity tolerance);
• Invasive and state-listed status (e.g., endangered, threatened)• Disturbed sites 1st three PC’s accounted for 68% variation;• No correlation with watershed population density;• Pattern in vegetation groups with degree marsh disturbance;
• Shift from obligate wetland plants with high salinity tolerance to less salinity tolerant species at disturbed sites;
• Invasive species more prevalent at disturbed sites.
High
Low
None
Medium
Salinity tolerance of obligate
wetland plants
Invasive plants
Increasing marsh disturbance
Vegetation – Possible metrics to detect change
• Vegetation groups:• Wetland status (e.g., obligate, facultative);• Salinity tolerance;• Invasive species;
• Pattern with degree of marsh disturbance.
John H. Chafee National Wildlife Refuge, RI
-8-6-4-20246810
-4 -2 0 2 4 6 8 10
Nekton Community – all species vs common species
• All species • Only 16% of the variability explained by 1st three principal components;• PC’s loaded primarily on rare species;
• 17 Common species, species present at >10% of marshes• Accounted for 34% of total variability;• No separation of sites;
• Conclusion: Individual species would not be a good indicator to detect change in nekton communities.
1st Principal Component
2nd P
rinci
pal C
ompo
nent
Reference sitesDisturbed sites
Common Nekton Species
-3
-2
-1
0
1
2
3
4
5
6
7
-3 -2 -1 0 1 2
Nekton Groups – based on life history characteristics
Reference sitesDisturbed sites
1st Principal Component
3RD P
rinci
pal C
ompo
nent
Grass Shrimp Dominated (Palaemonetes spp.) Resident Fish
Dominated
Transient Fishpresence
• Nekton Groups: Fish and crustaceans grouped by resident or transient species;
• 85% of variation explained by 1st three PCs;• PC1 showed negative correlation between resident fish & shrimp; • PC3 loaded on transient fish;
• Can use PC1 & PC3 to distinguish between sites with high proportions of resident fish, resident shrimp, & transient fish.
7
8
9
10
11
-3 -2 -1 0 1 2
Watershed Population Density
Nat
ural
log
of to
tal p
opul
atio
n
• First PC correlated with watershed population density for reference sites• First PC correlated with watershed population density;
• Sites in watersheds with higher population densities tended to be dominated by grass shrimp & sites in less dense watershed tended to be fish dominated.
1st Principal Component
Shrimp dominated
communities
Fish dominated communities
Pearson correlation coefficient: -0.64, p<0.0001, adj. r2: 0.40
-2
-1
0
1
2
3
4
-3 -2 -1 0 1 2
Nekton Groups & Population Density
1st Principal Component
3rd P
rinci
pal C
ompo
nent
Population of Surrounding watershed
Resident Shrimp(Palaemonidae spp.)
Resident fish(killifish spp.)
Transient fish(e.g., flounder,
eel)
1,000 to 4,9995,000 to 9,99910,000 to 19,99920,000 & above
Small watersheds surrounded by dense populations inNarragansett Bay, RI
• Sites in watersheds with high population densities were dominated by Palaemonidae grass shrimp;• As population density decreases, resident fish increase in abundance;• Transient fish are associated with lower population density watersheds sites.
Nekton– Possible metrics to detect change
• Nekton groups life history characteristics:• Resident fish;• Resident shrimp (Palaemonidae species);• Transient fish;
• Pattern with watershed population density.
•Although not strongly correlated in our analyses, these parameters are important:• Exotic species presence (exotic species can threaten native species);• Total species richness (generally, low richness is a sign of a degraded
community).
Striped killifish (resident fish) Young-of-the-year winter flounder (transient fish)
Grass shrimp Palaemonetes spp. (resident shrimp)
Estimate of Condition – Salt Marsh Vegetation
Vegetation Metric Condition Estimator (% cover)Good Moderate Poor
High salinity tolerant, wetland obligate >55% 40-55% <40%Low salinity tolerant, wetland obligate <2% 2-20% >20%Invasive plants <1% 1-20% >20%
0
10
20
30
40
50
60
70
80 Percent cover of vegetation metrics
Perc
ent C
over
Reference & OMWM marshes Restricted & Restoring marshes
Data graphed are:• 50th percentile (point) • 75th & 25% percentiles (box)• 90th & 10th percentiles (bars)
Data graphed are:• 50th percentile (point) • 75th & 25% percentiles (box)• 90th & 10th percentiles (bars)
Resident fish
Resident shrimp
Transient fish
Resident fish
Resident shrimp
Transient fish
0
0.2
0.4
0.6
0.8
1 Nekton metrics from reference marshes
Prop
ortio
n of
cat
ch
Low population watersheds High population watersheds
Resident fish
Resident shrimp
Transientfish
Resident fish
Resident shrimp
Transientfish
Estimate of Condition – Salt Marsh Nekton
Nekton Metric Condition EstimatorGood Moderate Poor
Resident fish (percent of catch) >60% 30-60% <30%
Resident shrimp (percent of catch) <15% 15-50% >50%
Transient fish (percent of catch) >15% 4-15% <4%
Exotic/Invasive species absent - present
Species richness ≥10 6-9 ≤5Breakpoints based on quartile values for reference marshes in watersheds with differing population densities.
Application to Resource ManagementAssign scores to each condition estimator, with high scores for better or more desired condition
• Good = 5, Moderate = 3, Poor = 1; • Presence / absence: absent=5; present = 1;• Sum the scores to obtain an overall value.• Set a range based on number of metrics,example for 3 metrics:
• Good: 12-15• Moderate: 7-11• Poor: 3-6
Assateague NS
Metric Condition Estimate
Score
Metric A Poor 1
Metric B Good 5
Metric C Moderate 3
Total score Moderate 9
Application to Resource Management
Vegetation data from Sachuest restricted & reference marshesVegetation metric Sachuest restricted
(before restoration, 1996)Sachuest restricted
(after restoration, 2002)Sachuest reference
(2002)
High salinity plants 38% (poor, 1) 77% (good, 5) 70% (good, 5)
Low salinity plants 25% (poor, 1) 25% (poor, 1) 1% (good, 5)
Invasive plants Present (poor, 1) Present (poor, 1) Absent (good, 5)
Total score POOR (3) MODERATE (7) GOOD (15)
Condition criteria applied to vegetation data from Sachuest Point NWR, RIA tidally restricted, degraded marsh that was tidally restored in 1997
Condition criteria applied to nekton data from Jamaica Bay, GATE
Nekton data from Big Egg & Elders marshes, GATE
Nekton metric Big Egg 2005 Big Egg 2009 Elders Lower 2009
Resident fish 46% (moderate, 3) 81% (good, 5) 39% (moderate, 3)
Resident shrimp 48% (moderate, 3) 7% (good, 5) 5% (good, 5)
Transient fish 1% (poor, 1) 8% (moderate, 3) 51% (good, 5)
Exotic species Present (1) Present (1) Absent (5)
Species richness 8 (moderate, 3) 10 (good, 5) 5 (moderate, 3)
Total score POOR (11) MODERATE (19) GOOD (21)
Conclusions
Cape Cod NS, MA
• Estimates were derived from a large regional multi-year database with a gradient of marsh disturbance from reference to severely degraded condition.
• Able to estimate condition criteria for salt marsh vegetation and nekton using key indicator metrics.
• Metrics were able to detect change in both vegetation and nekton communities when applied to new data.
Nekton Guilds
Salt Marsh Residents
Fish
Shrimp
Crabs
Salt Marsh Transients
Fish Crustaceans KillifishMinnows
Grass shrimp
Green crabsFiddler crabs
EelFlounder
Silversides
Blue crabsBay shrimp
Guilds based on life history characteristics associated with salt marshesResidents: salt marshes are critical to their life historyTransients: found in marshes, but not critical to their life history
Vegetation Groups - Surrounding Land use (5 km radius
• Canonical Correspondence Analyses (CCA);• Relationship among vegetation guilds and surrounding land use;• Vegetation groups from disturbed marshes were correlated with land use;
• Disturbed sites: 1st three CCA axes accounted for 67% of the variation (reference sites: 47% of variation);
• Higher proportions of invasive plants and low salinity tolerant species associated with developed lands.
-3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
Axis 1
Axis
2
Developed land (high, medium, low intensity) & open developed land
Wetlands & deciduous forest
Mixed forest & shrubWater & barren land
High salinity
Low salinity
Medium salinity
Salinity tolerance of wetland
obligate plants
Invasive plants
Disturbed sites
-1.5 -1 -0.5 0 0.5 1 1.5 2
-1
-0.5
0
0.5
1
1.5
2
Axis 1
Axis
2Nekton Groups – Surrounding Land use (5 km radius)
• Canonical Correspondence Analyses (CCA).• Nekton guilds from reference marshes were correlated with land use;o 1st two CCA axes accounted for 43% of the variation;o Higher proportions of resident shrimp (grass shrimp) were associated with
developed lands;o Higher proportions resident & transient fish were associated undeveloped lands.
Developed land (high, medium, low intensity & open)
Wetlands & mixed forest
Water & barren land, herbaceous vegetation
Crop land, shrub, deciduous forest
Resident fish
Resident shrimp
Transient fish
Nekton guilds
Resident crabs
sites
0
0.2
0.4
0.6
0.8
1
Can we estimate thresholds of change?Maybe for nekton….
Transient fish Resident fish Resident shrimp (Palaemonetes spp.)
Pro
porti
on o
f cat
ch
Watershed Population
If low population sites represent a more “natural” or “desired target” condition:>60% resident fish~15% transient fish~15% resident shrimp
If high population sites represent a more “stressed” condition :10% resident fish4% transient fish85% resident shrimp
0
0.2
0.4
0.6
0.8
1
<1000 1000-4000 4000-7000 7000-15000
Low pop Impacted sites, n=75
What is happening at impacted sites?At lower population watersheds (<15,000 people) shrimp are replacing the fish at the impacted marshes
Resident fish reference line: 68%
Shrimp & transient fish reference line: ~15%
Pro
porti
on o
f cat
ch
Watershed Population size
Impacted sites, n=75