KevCaz's Website

Do ecological interactions impact geographic distributions of species?

name: title
 layout: true
 class: center, middle, style1
 ---
 # Do ecological interactions impact geographic distributions of species?
 <hr />
 
 ## 2016 ESA Annual Meeting, Fort Lauderdale
 ### Wednesday, August 10
 
 ![:scale 60%](./img/bg.png)

### Kévin Cazelles     [http://kevincazelles.fr](http://kevincazelles.fr)    [](https://github.com/KevCaz)
 ### Nicolas Mouquet, David Mouillot & Dominique Gravel

---
 name: style2
 layout: true
 class: center, middle, style2
 ---
 count: false
 # CONTEXT
 <hr/>
 ### Biogeography and ecological interactions

---
    name: style1
    layout: true
    class: style1

---
 # Geographic distribution
 <hr/> 
 .center[![:scale 105%](./fig/fig1.png)]

???
    Espèces et sites

---
 # Four categories of factors shaping distributions
 <hr/> 
 ## 1. Climatic + edaphic
 --

## 2. Dispersion (landscape structure / dispersion capacities)
 --

## 3. Ecological interactions
 --

## 4. Evolution

???
 4 = dimension temporelle large échelle biogéographique historique
 je n'aborderais pas les questions reltives à l'évolution
 ---
 # 1. Climatic
 <hr/> 
 .center[![:scale 100%](./fig/fig3.png)]

???
    On fait une inférence non pas sur l'espace mais sur les facteurs climatiques.
    On change d'espace ; classiquement température / préciptation / nature des sols

---
 # 2. Dispersion
 <hr/> 
 .center[![:scale 87%](./fig/fig4.png)]

---
 # Species Distribution Models
 <hr/> 
 .center[![:scale 105%](./fig/fig5.png)]

???
    L'overlap mais que dit-il ?

---
 # 3. Ecological interactions
 <hr/>

## - B feeds on A
    --

.center[![:scale 80%](./fig/fig6b.png)]

--
    ## - Extinction probabilities impacted

---
 # A matter of spatial scale?
 <hr/> 
 .center[![:scale 87%](./fig/fig6.png)]

.right[[McGill, B. J. (2010). *Science*.](http://doi.org/10.1126/science.1188528)]

---
 # A matter of spatial scale?
 <hr/> 
 .center[![:scale 87%](./fig/fig6c.png)]

.right[[McGill, B. J. (2010). *Science*.](http://doi.org/10.1126/science.1188528)]
 .right[[Gotelli, N. J. (2010). *PNAS*.](http://www.pnas.org/cgi/doi/10.1073/pnas.0914089107)]

---
 # Ecological information within static occurrence data?
 <hr/>

--
    ## 1971: Brown, J. H. [Mechanisms of Competitive Exclusion Between Two Species of Chipmunks](http://doi.org/10.2307/1934589). Ecology.
    ### `$\rightarrow$` 2 chipmunks species + behavior + tree density = exclusion
    --

## 2014: Araújo, M. B., & Rozenfeld, A. [The geographic scaling of biotic interactions](http://doi.org/10.1111/j.1600-0587.2013.00643.x). Ecography.
 ### `$\rightarrow$` positive interactions easier to detect when scalling up

## 2016: Do ecological interactions impact geographic distributions of species?

---
 name: style2
 layout: true
 class: center, middle, style2
 ---
 count: false
 # THEORY
 <hr />
 ### Interactions and co-occurrence

---
    name: style1
    layout: true
    class: style1

---
 # Co-occurrence
 <hr/>

## - Observed co-occurrence : `$P(X_{A},X_{B})$`
    --

## - Independent co-occurrence : `$P(X_{A})P(X_{B})$`
    --

##         `$P(X_{A},X_{B})$` vs. `$P(X_{A})P(X_{B})$`

--
 
 ### Historically, species in competition [Diamond, 1975]()
 ### `$\rightarrow$` *Checkerboard* distribution
 ### `$\rightarrow$` *C-score*

???
    Si de telle manière alors elles intéragissent.

---
 # *Checkerboard* distribution
 <hr/> 
 .center[![:scale 60%](./fig/fig12.png)]

???
    $C_{ij} = (r_i- S_{ij}) (r_j - S_{ij})$
    C-score but for competition only

---
 # Z-score (standardized co-occurrence)
 <hr/>

# $$\frac{\text{"observed co-occurence"}-\mu}{\sigma}$$
    --

## `$\mu \rightarrow$` expected co-occurrence
    ## `$\sigma \rightarrow$` standard deviation associated
    --

## Scenarios
    --

###    1. Sites are identical (Homogeneous, [Veech 2013](http://doi.wiley.com/10.1111/j.1466-8238.2012.00789.x))
    --

###    2. Environmental conditions are taken into account (SDM)
    --

###        a. Generalised Linear Model (GLM)
    --

###        b. Random Forest (RF)

---
 # Shorthest path
 <hr/> 
 .center[![:scale 100%](./fig/fig14a.png)]

---
 # Shorthest path
 <hr/> 
 .center[![:scale 100%](./fig/fig14b.png)]

---
 # Shorthest path
 <hr/> 
 .center[![:scale 100%](./fig/fig14c.png)]

---
 # Shortest path + co-occurrence
 <hr/>

## 1. Distributions + network
    ### `$\rightarrow$` Nice datasets
    --

## 2. SDM for all species
    ### `$\rightarrow$` Integrating physiological requirements
    --

## 3. Network topology
    ### `$\rightarrow$` Shortest path for all pairs and mean degree in
    --

## 4. Co-occurrence for all pairs (3 scenarios)
    ### `$\rightarrow$` Look at distribution of Z-score

---
 # Expectations
 <hr/>
 .center[![:scale 95%](./fig/fig17.png)]

.right[[Cazelles K., *et al.* (2016) *Journal of Theoretical Ecology*.](http://doi.org/10.1111/ecog.01714)]

---
 # Expectations
 <hr/>
 .center[![:scale 95%](./fig/fig18.png)]

.right[[Cazelles K., *et al.* (2016) *Journal of Theoretical Ecology*.](http://doi.org/10.1111/ecog.01714)]

---
    name: style2
    layout: true
    class: center, middle, style2

---
 count: false
 # APPLICATION
 <hr />
 ### Real datasets

---
    name: style1
    layout: true
    class: style1

---
 # Salix / Herbivores / Parasitoids (>150 species)
 <hr/>
 .center[![:scale 60%](./fig/map_salix.jpeg)]

---
 # 10 Hummingbirds / 52 host plants on 32 islands
 <hr/>
 .center[![:scale 80%](./fig/map_hbirds.jpeg)]

.right[[Martín González., *et al.* (2015). *Global Ecology and Biogeography*](http://doi.org/10.1111/geb.12355)]

---
 # Micro-organisms community of the Pitcher plants leafs (>50 species)
 <hr/> 
 .center[![:scale 100%](./fig/map_pitcher.jpeg)]

.right[[Baiser, B., *et al.* (2012). *Global Ecology and Biogeography*](http://doi.org/10.1111/j.1466-8238.2011.00705.x)]

---
    # Salix / Herbivores

.center[![:scale 100%](./fig/figR1.png)]

---
    # Herbivores / Parasitoids

.center[![:scale 100%](./fig/figR2.png)]

---
    # Hummingbirds / host plants

.center[![:scale 100%](./fig/figR3.png)]

---
    # Pitcher plants leafs community

.center[![:scale 100%](./fig/figR4.png)]

---
 name: style2
 layout: true
 class: center, middle, style2
 ---
 # CONCLUSION
 <hr />
 ### Integrating interaction in biogeography

---
    name: style1
    layout: true
    class: style1

---
 # So, do ecological interactions impact geographic distributions of species?
 <hr/>

# YES

# Can we detect them is static occurrence data?

## `$\rightarrow$` it may be a matter of network topology.

--

## 1. `$\rightarrow$` in networks with high connectance, species can be considered as independent.

## 2. `$\rightarrow$` species separated by more than one link can be considered as independent.

???
    I feel like everyone feels so but the problem concerns predictions

---
 # Are our predictions reliable?
 <hr/> 
 .center[![:scale 100%](./fig/fig7.png)]

---
 # Are our predictions reliable?
 <hr/>

## `$\rightarrow$` as long as interactions are abundant enough

### `$\rightarrow$` if species assemblages move together

## `$\rightarrow$`How likely it is?
 ### Not very much I guess.

---
 # Toward a more integrative biogeography
 <hr/>

--
    ### - We should not assume *a priori* that species are independent units

--

### - Not only a spatial scale matter ...

--
    ### ... but also a matter of relevant biological unit for biogeography.

--
    .center[![:scale 75%](./fig/fig25.png)]

---
 # Toward a more integrative biogeography
 <hr/>

### - We should not assume *a priori* that species are independent units

### - Not only a spatial scale matter ...

### ... but also a matter of relevant biological unit for biogeography.

.center[![:scale 75%](./fig/fig26.png)]

---
    name: style2
    layout: true
    class: style2

---
    # ACKNOWLEDGMENT

<hr />

## People

- Tomas Roslin
    - Benjamin Baiser
    - Bo Dalsgaard
    - Louise J. Lehmann
    - Allan Timmermann
    - Ana M Martín González
    - Andrea Baquer

## Funding agencies

### - FRQNT
    ### - Frontenac

---
 name: style2
 layout: true
 class: center, middle, style2
 ---
 # THANK YOU FOR YOUR ATTENTION
 <hr/>
 ### Any question?

---
    name: style1
    layout: true
    class: style1

---

---
 # Co-occurrence and shortest path
 <hr/>
 .center[![:scale 100%](./fig/fig15.png)]

.right[[Cazelles K., *et al.* (2016) *Journal of Theoretical Ecology*.](http://doi.org/10.1111/ecog.01714)]

---

KevCaz's Website

Home

Blog

Miscelleneous

KevCaz's Website

Menu