COREIDS

Predicting community resilience to invasions from diversity and network structure

2014 – 2017

Project leaders: Patrice DAVID & François MASSOL

Post-docs: Grégory MOLLOT & Jelena PANTEL

Rabbit

Orycolagus cuniculus

Water lily

Eicchornia crassipes

Nile perch

Lates niloticus

Zebra mussel

Dreissena polymorpha

A worldwide revolution in

ecosystems and a major

component of global

change

Concerns all kinds of

organisms and habitats

Large impacts on

biodiversity but largely

unpredictable

biodiversity.eu

ropa.eu

Biological invasions

Networks of species interactions

Impacts of invasions

depend on interactions

with resident species

Interactions are organized

in networks (e.g. food

webs)

What properties make

networks sensitive to

invasion?

COREIDS project summary

• Assess the effects of invaders on biodiversity

• Understand which food webs are more likely

to be invaded

• Predict the effects of invaders based on

invaded food web and invader properties

• Reveal perturbation synergies between

anthropogenic pressures and invasions

The COREIDS group

A group of scientists

working on invasions

and/or networks in

France, Canada, New

Zealand

Field biologists,

experimental biologists,

statisticians, theoreticians

Syntheses, meta-analysis,

modelling

The COREIDS group

The COREIDS timeline01/0

9/2

014

2015 2016 20172014

meetings

final re

port

G. Mollot

J. Pantel

mid

repo

rtAdvances in Ecological Research issues 56 & 57

N. Poulain

The COREIDS meetings

Four meetings (Jan 2015, Sept 2015, March 2016 & Oct 2016)

At the CESAB

9-13 persons per meeting (total = 19)

Meeting #1: presentations, general discussions

Meeting #2: preparation of AER 56-57, tentative chapters

Meeting #3: update & work on chapters (small groups)

Meeting #4: updates on theoretical projects, presentations

Papers: 1. Reviews

Papers: 2. Data-based

Papers: 3. Models

Deliverables: wrap-up

Various reviews

Meta-analysis

Theoretical models

Statistical toolbox To do

Four types of local impacts of invaders on food webs

I

1. Top-down

Invasive

predator

Resident

prey

Euglandina rosea

(predatory “wolf” snail)

Endemic

Partula spp

(tree

snails)

Direct predation is responsible for most

species extinctions following invasions

(Moorea island)

R

Four types of local impacts of invaders on food webs

I

R1

2. Exploitativecompetition

R2

Invasive

species

Resident

prey

Resident

competitor

Locally

cultivated

fruit

(La Réunion island)

Endemic

Ceratitis catoiriiC. Capitata

Introd 1939

Four types of local impacts of invaders on food webs

I

R1

2. Exploitativecompetition

R2

Invasive

species

Resident

prey

Resident

competitor

Locally

cultivated

fruit

(La Réunion island)

Endemic

Ceratitis catoirii C. Capitata

Introd 1939C. rosa

Introd 1955

Four types of local impacts of invaders on food webs

I

R1

2. Exploitativecompetition

R2

Invasive

species

Resident

prey

Resident

competitorEndemic

Ceratitis catoirii

C. Capitata

Introd 1939

Locally

cultivated

fruit

C. rosa

Introd 1955

Bactrocera zonata

Introd 1991

- Invasive competitors often arrive in successive series

- Resident species are reduced or displaced more often than they go extinct

(La Réunion island)

Four types of local impacts of invaders on food webs

R

3. Bottom-up

- Resident predators may limit populations of invasive species (biotic resistance)

- Invasive species usually have positive effects on their predators

Invasive Gobies

Neogobius melanostomus

Endemic water snake

Nerodia sipedon insularum

(Lake Erie)

I

Invasive

prey

Resident

predator

Four types of local impacts of invaders on food webs

3. Bottom-up

- Invasive species usually have positive effects on their predators

- … but these benefits are sometimes outweighed by competition with more profitable

resident prey

- This situation is common with invasive plants e.g. Caulerpa taxifolia, Rubus alceifolius

>> herbivore diversity decreases

R2

Resident

competitor

Horned Lizard

Phrynosoma coronatum

Argentine ant

Linepithema humile

Native ant

Crematogaster californica

(California)

R

Invasive

prey

Resident

predator

I

Four types of local impacts of invaders on food webs

4. Apparent competition

Golden eagle

Aquila chrysaetos

Island fox

Urocyon littoralis

Feral pig

Sus scrofa

- Invasive species may benefit local generalist predators and indirectly affect local

species

(California channel islands)

R2

Resident

prey

R

Invasive

prey

Resident

predator

I

• ASYMMETRY: Invaders tend to

have more impact (often strong) on

natives than natives have on

invaders; they are fiercer

predators, better-defended prey,

and stronger competitors

The filter hypothesis

Predation

efficiency

coevolved

Not coevolved

Frequency

Example of predator introduction

• ASYMMETRY: Invaders tend to have more impact (often strong) on natives than natives have on invaders; they are fiercer predators, better-defended prey, and strongercompetitors

• LACK OF COEVOLUTION: Invadersand residents are evolutionary naiveto each other : their interactions may often be extreme

• FILTER: Only species that happen to be tolerant to resident speciessuccessfully invade

• Major outcomes

The filter hypothesis

• ASYMMETRY: Invaders tend to

have more impact (often strong) on

natives than natives have on

invaders; they are fiercer

predators, better-defended prey,

and stronger competitors

• LACK OF COEVOLUTION:

Invaders and residents are

evolutionary naive to each other :

their interactions may often be

extreme

• FILTER: Only species that happen

to be tolerant to resident species

successfully invade

• Major outcomes

Predation

efficiency

coevolved

Not coevolved

Mean

natives

Invasion

failsInvasion

succeeds

Mean

invaders

Frequency

Example of predator introduction

The filter hypothesis

Propagation of effects in the recipient network: attenuation or amplification?

I

Theoretical predictions

• Theoretical models: construction of model food webs by

addition of species and evolution of trophic traits (e.g. body

size) or classic models of food web simulation (niche model)

• Introduce virtual invaders, measure probability of success

and impact

• Diversity >> filling of niche space >> fewer invasions and

lower impacts

• Connectance: decrease probability of invasion, but potentially

more negative effects if invasion succeeds

Theoretical predictionsDistribution of robustness (=% of persistent species after invasion attempt) by

connectance

Theoretical predictionsDistribution of robustness (=% of persistent species after invasion attempt) by

connectance & by type of invader

All impacts

At least one species

lost

Theoretical predictionsExtinction cascades depend on food web topology

Empirical data: Three cases of far-reaching impact of species introduction

1) Trophic

cascades in lakes

Favored by the relatively low

diversity and high efficiency

of primary consumption of

phytoplankton

(unlike in many plant-based

terrestrial ecosystems)

http://www.lmvp.org

Empirical data: Three cases of far-reaching impact of species introduction

2) Disequilibrium

- Recent human modification:

eutrophication, elimination of top

predators, replacement of native

forest by agriculture or secondary

habitat >> low diversity, lots of

unexploited resource

- Evolutionary immaturity: remote

islands lacking particular groups

>> idem

http://www.lmvp.org

Guam island

Tree snake

Boiga irregularis

Empirical data: Three cases of far-reaching impact of species introduction

3) Ecosystem

engineering

- Large modification of habitat:

sessile filter-feeders; water

clarification, creation of habitats

- Attracts many new species,

limits many others, deeply

reorganizes energy and matter

fluxes in lake ecosystem

(pelagic >> benthic)http://ian.umces.edu

Zebra mussel

Great Lakes

Some invaders have more effect

Some ecosystems are more affected

Some regions are more affected

Lessons for ecosystem conservation

• Fight against causes of invasions rather than invasions

themselves: pollution, perturbation, predator removal

• Islands are hypersensitive due to past isolation: prevent

introductions, especially of generalist predators

• Eradication is usually costly and ineffective: resident

communities often change ecologically and evolutionarily after

invasions, there is often no way back. Look for mitigation

rather than eradication.

Thank you for your attention