Risk Assessment for New Zealand
Mudsnail (Potamopyrgus antipodarum)
in Canadian Waters
Thomas W. Therriault1, Andrea M. Weise2,
Graham E. Gillespie1, and Todd J. Morris3
Department of Fisheries & Oceans Canada
1) Pacific Biological Station, Nanaimo, BC
2) Institut Maurice-Lamontagne, Mont Joli, QC
3) Canadian Centre for Inland Waters, Burlington, ON
• New Zealand mudsnail (NZMS) was identified as species of concern by CEARA based on global invasion history.
• With reports from the Great Lakes and Pacific Coast, a formal risk assessment was deemed necessary to characterize the potential risk posed to Canadian waters by this species.
Introduction
• CEARA adopted DRAFT guidelines for conducting biological risk assessments in 2006. These guidelines were revised in 2008 but have not been finalized.
• The initial guidelines were developed based on the Canadian National Code on Introductions and Transfers of Aquatic Organisms while the most recent version follows a general invasion cycle.
General RA Framework
• Step 1: Determining the Probability of a Widespread Invasion.
• Step 2: Determining the Consequences (Impacts) of a Widespread Invasion.
• Step 3: Combining the Probability of a Widespread Invasion with the Consequences (Impacts) of a Widespread Invasion (Determining Risk)
General RA Framework
RAs for NIS
Survive
Arrive
Spread
Reproduce
Consequences
• Risk = Probability X Impact (Consequence)P1
P4
P3
P2
I1 I2 I3I4
Stages of InvasionEvent Probability
(P)
Impact (I) Impact Description
Arrive P1 I1 Species arrives but does
not survive
Survive P2 I2 Species survives but
does not establish
Establish P3 I3 A local population is
established
Spread P4 I4 Widespread invasion
Stages of InvasionOutcome Probability Impact Risk
Arrives (A), but no
survival
PA = Min(P1, 1 – P2) IA = I1 RA = (PA,IA)
Survives (S), but
no reproduction
PS = Min(P1,P2, 1 – P3) IS = I2 RS = (PS,IS)
Establishes (E)
locally, not
widespread
PE = Min(P1,P2, P3,1 – P4) IE = I3 RE = (PE,IE)
Widespread (W)
invasion
PW = Min(P1,P2, P3, P4) IW = I4 RW = (PW,IW)
Overall risk RTotal = Max(RA,RS,RE,RW)
The Risk Matrix
Impact /
Consequence
Very High
High
Moderate
Low
Very Low
Positive
Very Low Low Moderate High Very High
Probability of Invasion
• Temporal: 5 years
• Spatial
1. Laurentian Great Lakes/St. Lawrence estuary;
2. Pacific Freshwater;
3. Arctic Freshwater;
4. Gulf of Mexico Freshwater;
5. Hudson Bay Freshwater;
6. Atlantic Freshwater;
7. Pacific Coastal; and
8. Atlantic Coastal
Scope and Scale
Canadian Freshwater Drainages
• Risk was assessed for each unit based on:
– Literature (primary and gray)
– Environmental Niche Modeling (GARP)
– Expert Survey
• Vectors/Pathways
• Impacts
– Input from CSAS Review Meeting
Additional Information
Biological Review and Current
Distribution of
New Zealand Mudsnail
Predicting the Potential
Distribution of New Zealand
Mudsnail in Canada
GARP Model
• Genetic Algorithm for Rule-set Prediction
• Used geo-referenced environmental data
– 0.5 degree grid
– Terrestrial and climate surrogates for freshwater data
• NZMS current North American distribution
– 0.5 degree grid
– Reduces presence records from 1,865 to 224
Environmental Data
• Elevation (m)
• Flow accumulation (km2)
• Ground frost frequency
(d)
• Maximum annual
temperature (1961-1990,
C)
• Mean annual temperature
(1961-1990, C)
• Minimum annual
temperature (1961-1990,
C)
• Precipitation (mm)
• Slope (m)
• Solar radiation (W/m2)
• Topographic index
• Wet-day frequency (d)
• Surficial geology
Potential Canadian Distribution
Eastern Detail
Western Detail
Variable Hierarchical partitioning (%)
Maximum annual air temperature 48.1
Mean annual air temperature 23.4
Minimum annual air temperature 13.8
Radiation 8.5
Flow accumulation 2.9
Topographic index 2.5
Slope 0.6
Precipitation 0.1
Wet day index 0.1
Partitioning of Variables
Using an Expert Survey to
Inform the Risk Assessment of
New Zealand Mud Snail in
Canada
• Survey primarily targeted individuals familiar with NZMS but also sent to experts on invasive species/gastropods
• Of 125 individuals targeted, 70 responded
• Identified the relative importance of 2 primary and 30 secondary invasion vectors
• Identified impacts to 8 ecological endpoints
• Uncertainty also characterized
Expert Survey Overview
Vector ImportanceVector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Primary Vector
Commercial movement of contaminated aquaculture products (live fish and eggs)
13(3)
12(4)
10(14)
5(14)
1(6)
41(41)
Ballast water – foreign port
7(6)
16(4)
5(13)
8(13)
5(6)
41(42)
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Secondary Vectors
Attachment to gear (i.e. waders, boots, etc.) (RA)
17(1)
17(3)
6(6)
1(16)
0(11)
41(37)
Indirect attachment as fouling through mud, macrophytes and algae (RWU)
15(2)
15(3)
9(11)
2(14)
0(8)
41(38)
Rheotactic movement (ND)
13(2)
7(4)
9(12)
7(9)
1(7)
37(34)
Natural drift (ND) 10(3)
8(5)
8(14)
10(8)
2(5)
38(35)
Vector Importance
Vector types: Natural Dispersal (ND), Transport by Secondary Organisms (TSO),
Natural Resources Activities (NRA), Fish Hatcheries (FH), Recreational Anglers
(RA) and Recreational Water Users (RWU).
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Contaminated hatchery transplants - live passage through fish digestive systems (FH)
6(3)
16(5)
11(10)
4(10)
1(6)
38(34)
Contamination of water supplies (FH)
5(1)
16(6)
14(10)
3(12)
0(5)
38(34)
Professional field research gear and clothing (NRA)
8(1)
15(3)
9(12)
4(12)
2(7)
38(35)
Waterway operations removing or transporting substrate material (i.e. dredging) (NRA)
4(4)
15(8)
12(11)
6(8)
1(3)
38(34)
Attachment to floating aquatic macrophytes and algae (TSO)
8(3)
14(3)
9(17)
7(10)
1(5)
39(38)
Fish Stocking (FH) 8(1)
14(6)
10(13)
4(11)
1(4)
37(35)
Vector Importance
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Flood events causing dislodgement - natural or human caused (ND)
5(3)
14(6)
10(13)
8(9)
1(4)
38(35)
Within bird and fish gastrointestinal tracts (TSO)
3(4)
14(5)
12(13)
9(12)
2(3)
40(37)
Inadvertently distributed in live bait or when bait containers are discarded (RA)
3(3)
13(7)
9(16)
13(8)
0(0)
38(34)
Embedded in mud on anchors and other watercraft accessories (RWU)
10(3)
12(4)
10(12)
6(11)
0(5)
38(35)
Entrained in water lines or livewell/bilge of watercraft's (RWU)
3(2)
12(8)
10(11)
11(10)
1(3)
37(34)
Attachment to aquatic ornamental plants -wholesale/retail (TSO)
7(2)
11(7)
10(12)
10(10)
0(4)
38(35)
Vector Importance
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Freshwater tanks 3(6)
7(6)
16(10)
9(7)
1(4)
36(33)
Citizen and classroom monitoring groups (NRA)
6(2)
9(4)
15(15)
7(10)
1(4)
38(35)
Canal and ditch maintenance (NRA)
3(5)
11(7)
15(13)
7(6)
1(3)
37(34)
Passive transport by feet or fur of domestic livestock (TSO)
0(6)
5(9)
15(10)
13(8)
4(1)
37(34)
Firefighting machinery or equipment used to fight backcountry forest fires (NRA)
3(7)
5(10)
14(6)
13(9)
3(3)
38(35)
Direct attachment as fouling to watercraft/trailers (RWU)
6(3)
10(6)
13(14)
10(9)
1(4)
40(36)
Vector Importance
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Hikers, Backpackers, Horseback riders, Bicyclists (RWU)
0(4)
4(8)
14(14)
19(8)
2(1)
39(35)
Deliberate release 3(9)
2(5)
5(7)
16(10)
12(5)
38(36)
Movement of live/dead fish between watersheds (RA)
2(4)
9(7)
12(15)
16(8)
0(2)
39(36)
Volitional movement (ND)
4(3)
8(6)
7(6)
16(13)
1(5)
36(33)
Within bird bills, feathers or on legs (TSO)
2(6)
6(6)
13(12)
15(9)
1(2)
37(35)
Passive transport by semi-aquatic and terrestrial wildlife (TSO)
0(5)
8(10)
13(16)
15(3)
3(2)
39(36)
Vector Importance
Vector Importance
Vector of Introduction / Spread
Very High
High Moderate Low Very Low
N
Free-Floating juveniles on water surface (ND)
8(5)
6(5)
9(12)
13(9)
1(3)
37(34)
Ballast discharge -coastal shipping
3(4)
8(3)
9(13)
10(8)
7(7)
37(35)
Vector Importance
ImpactsImpact Level
Negative Positive
Ecological Endpoint Very High High Moderate Low Very Low N
Freshwater Biodiversity 6 14 12 5 3 1 41
Protected Areas /
Conservation Areas7 12 6 5 2 1 33
Aquaculture 6 5 8 5 6 1 31
Recreational Fisheries 4 7 12 6 7 0 36
Habitat 7 6 5 10 6 1 35
Estuarine Biodiversity 0 7 7 9 8 1 32
Marine Biodiversity 0 0 1 4 25 0 30
Commercial Fisheries 1 2 10 6 14 0 33
Step 1: Determining the
Probability of a Widespread
Invasion
Region P1 P2 P3 P4
Great Lakes / St Lawrence Very High
(Very Low)
Very High
(Very Low)
Very High
(Very Low)Very High
(Low)
Pacific Freshwater Very High
(Low)
Very High
(Low)
Very High
(Low)
Low
(Moderate)
Arctic Freshwater Low
(Very High)
Very Low
(Moderate)
Very Low
(Moderate)
Very Low
(Low)
Gulf of Mexico Freshwater High
(High)
Very High
(Low)
Very High
(Low)High(Low)
Hudson Bay Freshwater High
(Low)
Very High
(Low)
Very High
(Low)
Moderate
(Moderate)
Atlantic Freshwater Very High
(Very Low)
Very High
(Low)
Very High
(Low)
Moderate
(Moderate)
Pacific Coastal Very High
(Very Low)
Very High
(Very Low)
Very High
(Very Low)
Low
(High)
Atlantic Coastal Low
(Very High)
Very High
(Low)
Very High
(Low)
Low
(Very High)
Ranks (Uncertainty)
Region
PA - arrival
without
survival
PS - survival
without
establishment
PE
establishment
without spread
PW
widespread
invasion
Great Lakes / St Lawrence Very Low Very Low Very Low Very High
Pacific Freshwater Very Low Very Low High Low
Arctic Freshwater Low Very Low Very Low Very Low
Gulf of Mexico Freshwater Very Low Very Low Low High
Hudson Bay Freshwater Very Low Very Low Moderate Moderate
Atlantic Freshwater Very Low Very Low Moderate Moderate
Pacific Coastal Very Low Very Low High Low
Atlantic Coastal Very Low Very Low Low Low
Overall Probabilities
Step 2: Determining the
Consequences (Impacts) of a
Widespread Invasion
Region Biodiversity Habitat Aquaculture Commercial
Fisheries
Recreational
Fisheries
Great Lakes / St Lawrence Low
(High)
Low
(High)
Very Low
(Moderate)
Low
(High)
Low
(High)
Pacific Freshwater Moderate
(Very High)
Low
(High)
Very Low
(High)
Very Low
(High)
Low
(High)
Arctic Freshwater Moderate
(Very High)
Low
(High)
Very Low
(High)
Very Low
(High)
Very Low
(High)
Gulf of Mexico Freshwater Moderate
(Very High)
Low
(High)
Very Low
(High)
N/A Low
(High)
Hudson Bay Freshwater Moderate
(Very High)
Low
(High)
Very Low
(High)
Low
(High)
Low
(High)
Atlantic Freshwater Moderate
(Very High)
Low
(High)
Very Low
(High)
Very Low
(High)
Low
(High)
Pacific Coastal Very Low
(High)
Very Low
(High)
Very Low
(High)
Very Low
(High)
Very Low
(High)
Atlantic Coastal Very Low
(High)
Very Low
(High)
Very Low
(High)
Very Low
(High)
Very Low
(High)
Rank of Potential Impacts
Step 3: Determining the
Overall Risk of a Widespread
Invasion
The Risk Matrix
Impact /
Consequence
Very High
High
Moderate
Low
Very Low
Positive
Very Low Low Moderate High Very High
Probability of Invasion
Region Biodiversity Habitat Aquaculture Commercial
Fisheries
Recreational
Fisheries
Great Lakes / St Lawrence Low
(High)
Low
(High)
Low
(Moderate)
Low
(High)
Low
(High)
Pacific Freshwater Low
(Very High)
Low
(High)
Low
(High)
Low
(High)
Low
(High)
Arctic Freshwater Low
(Very High)
Low
(High)
Low
(High)
Low
(High)
Low
(High)
Gulf of Mexico Freshwater Moderate
(Very High)
Low
(Very High)
Low
(High)N/A
Low
(High)
Hudson Bay Freshwater Moderate
(Very High)
Low
(High)
Low
(High)
Low
(High)
Low
(High)
Atlantic Freshwater Moderate
(Very High)
Low
(High)
Low
(High)
Low
(High)
Low
(High)
Pacific Coastal Low
(High)
Low
(High)
Low
(High)
Low
(High)
Low
(High)
Atlantic Coastal Low
(Very High)
Low
(Very High)
Low
(Very High)
Low
(Very High)
Low
(Very High)
Overall Risk Assessment
Conclusions
Conclusions
• Overall risk of NZMS to Canadian ecosystems was designated low to moderate
– Moderate risk to Freshwater Biodiversity in Gulf of Mexico, Hudson Bay and Atlantic drainages
– Low for all other ecological/geographical combinations
• Risk at smaller scales could be significantly higher
Conclusions
• Vectors have high levels of uncertainty
– More work needed to decrease uncertainty
• Difficult to predict impacts on Canadian
ecosystems
– More work needed to understand NZMS biology and
ecology
• Current distribution uncertain
– Education and outreach for better monitoring of NZMS
distribution in Canada
Conclusions
• Niche modeling can be improved
– Need to update as new occurrences are reported
– Need to explore other environmental layers
– Modeling at smaller scales may be appropriate
for management advice
Conclusions
• This assessment represents a starting point
– This biological assessment can inform overall
assessments that incorporate socio-economic
considerations, risk management and risk
communication
– Should be revisited as new data become
available and information gaps are filled
Documents
• Available at the Canadian Science Advisory
Secretariat website
– http://www.isdm-gdsi.gc.ca/csas-
sccs/applications/publications/index-eng.asp
– CSAS Proceedings Series 2010/023
– CSAS Research Document 2010/108
– CSAS Science Advisory Report 2010/065
Thank you!