# NetworkExtinction: Extinction Simulation in Food Webs

Derek Corcoran, M.Isidora Avila Thieme, Pablo A. Marquet, Sergio A. Navarrete y Fernanda S. Valdovinos
"26/03, 2018"

### Indexes

• Primary removals: It occurs when the researcher intentionally removes one species, simulating a single extinction.
• Secondary extinctions: A secondary extinction occurs when a non basal species loses all of its prey items due to the removal of another species.
• Total extinctions: The sum of primary removal and secondary extinctions in one simulation. .

### Mostconnected

library(NetworkExtinction)
data("net")
Mostconnected(Network = net)

6 9 7 0.09 0.78 1 1 1 1 2
7 7 4 0.08 0.57 0 2 1 2 3
5 6 2 0.06 0.33 1 3 2 3 5
2 4 0 0.00 0.00 1 4 3 4 7
• Recursive extinction from most to least connected

### Plotting Mostconnected

history <- Mostconnected(Network = net)
ExtinctionPlot(History = history, Variable = "AccSecondaryExtinction")


### Extinctions using a customized order

data("net")
ExtinctionOrder(Network = net, Order = c(2,4,7))

Spp nodesS linksS Conectance Secondary_extinctions AccSecondaryExtinction NumExt TotalExt
2 9 8 0.10 1 1 1 2
4 7 5 0.10 1 2 2 4
7 5 3 0.12 0 2 3 5

### RandomExtinctions

data("net")
RandomExtinctions(Network= net, nsim= 50)

NumExt SdAccSecondaryExtinction AccSecondaryExtinction
1 0.43 0.24
2 0.61 0.56
3 0.82 1.02
4 0.88 1.44
5 1.01 1.86
6 1.22 2.20
7 1.14 2.16
8 1.11 2.35
9 1.25 2.29

### CompareExtinctions

• Comparison of Null hypothesis with other extinction histories
data("net")
History <- ExtinctionOrder(Network = net, Order = c(1,2,3,4,5,6,7,8,9,10))
NullHyp <- RandomExtinctions(Network = net, nsim = 100)
Comparison <- CompareExtinctions(Nullmodel = NullHyp, Hypothesis = History)


### CompareExtinctions (Continued)

Comparison$Test   Pearson's Chi-squared test data: Hypothesis$DF$AccSecondaryExtinction and Nullmodel$sims$AccSecondaryExtinction[1:length(Hypothesis$DF\$AccSecondaryExtinction)]
X-squared = 20, df = 16, p-value = 0.2202

• Kolmogorov-Smirnov?
• Otras alternativas?

### degree_distribution

data("chilean_intertidal")
degree_distribution(chilean_intertidal, name = "Test")


### degree_distribution

sigma isConv finTol logLik AIC BIC deviance df.residual model
0.09 TRUE 0 66.06 -128.12 -123.68 0.57 67 Exponential
0.22 TRUE 0 5.42 -6.84 -2.46 3.28 65 Power
0.48 TRUE 0 -45.32 94.64 99.02 15.26 65 truncated