Lecture 21: Risk Assessment

Question 1

Risk?

Answer 1

prob of establishment x prob of impact

Question 2

What is a simple model to prioritize invasion threats?

Answer 2

Threat index: 
Colonization potential 
Impact potential 
Feasibility of control
**if colonization potential is zero, threat index = 0

Question 3

What is colonization potential?

Answer 3

• probability of introduction ( propagule pressure) and probability of establishment (biological traits)

Question 4

What is impact potential?

Answer 4

the probability of altering ecosystem processes or of the removal of native species

• can be predicted from a species invasion history
• a function of characteristics of both the invader and the recipient community

Question 5

What is feasibility of control?

Answer 5

• only once established
• high impact invaders are difficult to control are a high priority for PREVENTION
• lower feasibility gets higher scores

Question 6

How do you assess colonization potential?

Answer 6

• escape potential (proximity to natural water bodies, type of culture system)
• acclimatization potential (habitat suitability, reproductive potential, dispersal potential)

Question 7

How do you assess impact potential ?

Answer 7

vulnerability of the receiving system
ex endangered native species
- threat of the potential invader
L> habitat alteration, food web alteration, gene pool alteration, disease introduction

Question 8

Impact =?

Answer 8

outcome of the interaction of the invader and its environment

Impact = invaders abundance, invaders ecological function, composition of the invaded community

Question 9

How do we predict impacts given the context dependency of invasions?

Answer 9

• experiments + statistical synthesis (meta analysis and empirical modelling )
  L> SS = macroecological approach ( identify predictable patterns using data form many invaded sites)

Question 10

How to choose an empirical approach to predict impact?

Answer 10

1. Does the species have an impact history?
   L> Yes = identify patterns within the data
   L> No = Does it have an invasion history?
   A. Yes= use models of abundance
   B. Use impact history of a similar species

Question 11

Approach 1?

Answer 11

compare impacts at different invasion sites (consistency = predictability)
ex: Green crab preferentially eats molluscs regardless where it lives
ex: Carp consistently increases veg distribution as its biomass increases at multiple sites
ex: Zebra mussels consistently cause decreases in suspended particles, phytoplankton production, zooplankton biomass
and increases in transparency, macrophyte biomass ( increase with higher light access but eventually will reach a carrying capacity and will crash, stripping the water of oxygen = phyto decline), benthic invert density and waterfowl density

Question 12

Approach 2?

Answer 12

• build statistical relationships using data from multiple invaded sites

Question 13

If an invader doesn’t have an impact history but does have an invasion history we look at what to predict impact risk?

Answer 13

• abundance of invader
  ex: carp: higher biomass = lower veg biomass
  ex: higher zebra mussel biomass with higher calcium content

Question 14

Environmental matching hypothesis ?

Answer 14

impact of a species is inversely correlated with distance from its environmental optimum
ex: fouling of zebra mussels varies with calcium content

Question 15

What do we do if a species doesn’t have an impact history or invasion history ?

Answer 15

• compare it to a functionally similar species that does have an invasion history
  ex comparing golden mussel (no history) to zebra (has a history )
  **Does not guarantee they will behave the same though

Question 16

Composition of the receiving community?

Answer 16

• presence of predators
  ex: zebra mussel impact lowers in the presence of the round goby
• presence of an organism that facilitates invasion /environmental conditions
  ex bryozoan and Codium plus acidification of the ocean

Question 17

Summary?

Answer 17

1. impacts are context dependent but are not entirely unpredictable
2. Predictive models can be generated by synthesis of data from many sites
   L> impact history can reveal types, direction and magnitude of effects
   L> most variation may be related to a few key physical variables
3. Presently there are insufficient data to predict the impacts of most aquatic invaders
4. Impacts of species without impact histories can be estimated from abundance models
5. Impacts are a function of the properties of the invaded community
   L> generalizations cannot be made from a few case studies