Stano Pekár“Populační ekologie živočichů“
dN
= Nr
dt
Acarus Cheyletus
continuous model of Lotka & Volterra (1925-1928)
- continuous predation
- capture several prey items, functional response Type II
H .. density of prey P .. density of predators
r .. intrinsic rate of prey population m .. predator mortality rate
a .. predation rate b .. reproduction rate of predators
in the absence of predator, prey grows exponentially →
in the absence of prey, predator dies exponentially →
predation rate is linear function
of the number of prey .. aHP
each prey contributes identically
to the growth of predator .. bHP
rH
t
H
=
d
d
mP
t
P
−=
d
d
aHPrH
t
H
−=
d
d
mPbHP
t
P
−=
d
d
prey population would grow to infinity
→ neutral stability
do not converge, has no asymptotic
stability (trajectories are closed lines)
unstable system, amplitude of the cycles
is determined by initial numbers
POOR model
Zero isoclines:
for prey population:
for predator population:
H
P
prey isocline
predator isocline
0
0
d
d
=
t
P
aHPrH −=0
a
r
P =
mPbHP −=0
b
m
H =
0
d
d
=
t
H
timedensity
prey
predator
0
Analysis of the model
a
r
b
m
in the absence of the predator prey population reaches
carrying capacity K
Incorporation of density-dependence
for given parameter values: r = 3, m = 2, a = 0.1, b = 0.3, K = 10
HP
H
H
t
H
1.0
10
13
d
d
−





−= PHP
t
P
23.0
d
d
−=
aHP
K
H
rH
t
H
−





−= 1
d
d
mPbHP
t
P
−=
d
d
Zero isoclines:
for prey population:
if H = 0 (trivial solution) or if
for predator population: 0.3HP - 2P = 0
if P = 0 (trivial solution)
or if 0.3H - 2 = 0
gradient of prey isocline is negative
0
d
d
=
t
H
HP
H
H 1.0
10
130 −





−=
0
d
d
=
t
P
P = 30 - 3H
P
H
1.0
10
130 −





−=
H = 6.667
H
P
30
6.670 10
time
density
prey
predator
0
K
H
P
30
6.70 10
.. approaches stable equilibrium
functional response Type II:
rate of consumption by all predators:
Incorporation of functional response
for parameters: rH = 3, a = 0.1, Th = 2, K = 10
prey isocline: predator isocline:
h
a
aHT
aHT
H
+
=
1
h
a
aHT
aHP
T
PH
+
=
1
h
H
aHT
aHP
K
H
Hr
t
H
+
−





−=
1
1
d
d
0
d
d
=
t
H
21.01
1.0
10
130
H
HPH
H
+
−





−=
2
6.0630 HHP −+=
mPbHP
t
dP
−=
d
H = constant
.. damped oscillations
predator exploits prey
close to K
- isocline: H = 9
time
density
time
density
time
density
predator exploits
prey close to K/2
- isocline: H = 5
predator exploits
prey at low density
- isocline: H = 2
Rosenzweig & MacArthur (1963)
H
P
H
P
H
P
K
prey
predator
0 0 0
0 0 0K/2
K
Damped oscillations Sustained oscillations Extinction
K K
logistic model with carrying capacity proportional to H
k .. carrying capacity of the predator
rP = bH - m
Incorporation of predator’s carrying capacity
for parameters: rP = 2, k = 0.2
predator isocline:
prey isocline:
mPbHP
t
P
−=
d
d






−=
kH
P
Pr
t
P
P 1
d
d
0
d
d
=
t
P






−=
H
P
P
2.0
120
H = 5P
2
6.0630 HHP −+=
h
H
aHT
aHP
K
H
Hr
t
H
+
−





−=
1
1
d
d
H
P
K0
time
density
prey
predator
0
K
H
P
K0
.. quick approach to stable equilibrium
Zatypota Theridion
discrete model of Nicholson & Bailey (1935)
- discrete generations
- 1, .., several, or less than 1 host
- random host search and functional response Type III
- lay eggs in aggregation
Ht = number of hosts in time t
Ha = number of attacked hosts
λ = finite rate of increase of the host
Pt = number of parasitoids
c = conversion rate, no. of parasitoids for 1 host
)(1 att HHH −=+ λ
aat HcHP ==+1
parasitoid searches randomly
encounters (x) are random (Poisson distribution)
p0 = proportion of not encountered, µ .. mean number of encounters
Et = total number of encounters
a = searching efficiency (proportion of hosts encountered)
Et = a Ht Pt
proportion of encounters (1 or more times): p = (1– p0)
Incorporation of random search
x = 0, 1, 2, ...
!x
e
p
x
x
µ
µ −
=
µ−
= ep0
( )taP
ta eHH −
−= 1
t
t
t
aP
H
E
==µ taP
ep −
=0
)1( taP
ep −
−=
highly unstable model for all parameter values:
- equilibrium is possible but the slightest disturbance leads to divergent
oscillations (extinction of parasitoid)
taP
tt eHH −
+ = λ1
( )taP
tt eHP −
+ −= 11
time
density
H
P
0
0
)(1 att HHH −=+ λ
at HP =+1
exponential growth of hosts is replaced by logistic equation
H*.. new host carrying capacity
depends on parasitoids’ efficiency
- when a is low then q → 1
- when a is high then q → 0
density-dependence have
stabilising effect for moderate r and q
Stability boundaries
Incorporation of density-dependence
Beddington et al. (1975)
t
t
aP
K
H
tt eHH
−





−
+ =
1
1 λ
( )taP
tt eHP −
+ −= 11
K
H
q
*
=
Incorporation of the refuge
if hosts are distributed non-randomly in the space
Fixed number in refuge: H0 hosts are always protected
have strong stabilising effect
even for large r
Hassell & May (1973)
taP
tt eHHHH −
+ −+= )( 001 λλ
( )taP
tt eHHP −
+ −−= 1)( 01
distribution of encounters is not random but aggregated (negative
binomial distribution)
- proportion of hosts not encountered (p0):
where k = degree of aggregation
very stable model system if k ≤ 1
Stability boundaries:
a) k=∝, b) k=2, c) k=1, d) k=0
Incorporation of aggregated distribution
Hassell (1978)
k
tt
k
aP
K
H
tt eHH
−






+





−
+ =
11
1 λ














+−=
−
+
k
t
tt
k
aP
HP 111
k
t
k
aP
p
−






+= 10
You want to control population of mites. Before introduction of
predatory mites you want to simulate the predator-prey dynamic using
the following model:
Parameter estimates are obtained experimentally:
1. Rear prey population without predators. You find rH = 0.4 and K =
500.
2. Rear predators at constant prey densities. You find predators’
mortality d = 0.08 and conversion efficiency c = 0.8.
3. Perform functional response experiment. You find that a = 0.001 and
Th = 0.5.
How long it takes for the predatory mite to control mite pests. Initial
densities are 200 individuals of pests and 10 individuals of predators?
h
H
aHT
aHP
K
H
Hr
t
H
+
−





−=
1
1
d
d
dP
aHT
acHP
t
P
h
−
+
=
1d
d
predprey<-function(t,y,pa){
H<-y[1]
P<-y[2]
with(as.list(pa),{
dH.dt<-rH*H*(1-H/K)-a*H*P/(1+a*H*Th)
dP.dt<-a*c*H*P/(1+a*H*Th)-d*P
return(list(c(dH.dt,dP.dt)))})}
H<-200;P<-10
time<-seq(0,200,0.1)
pa<-c(rH=0.4,K=500,a=0.001,Th=0.5,c=0.8,d=0.08)
library(deSolve)
out<-data.frame(ode(c(H,P),time,predprey,pa))
matplot(time,out[,-1],type="l",lty=1:2,col=1)
legend("right",c("H","P"),lty=1:2)
Caterpillars increased their population density in flour to 50
individuals/100 kg. You observed that their λ = 3 and K = 800. You need
to control these pests using a parasitoid. You can choose from three
parasitoid species (A, B, C). The three species differ in the number of
eggs/host (c) and in their search efficiency (a):
1. Use the discrete Nicholson-Bailey host-parasitoid model with densitydependence.
Introduce a single parasitoid per 100 kg. Find which of the
three species will achieve the quickest control.
A B C
c 1 3 2
a 0.003 0.1 0.005
time<-20
HP<-data.frame(H<-numeric(time),P<-numeric(time))
a=0.003;L=3;K=800;c=1
HP[1,]<-c(50,1)
for (t in 1:20) HP[t+1,]<-{
H<-L*HP[t,1]*exp((K-HP[t,1])/K-a*HP[t,2])
P<-c*HP[t,1]*(1-exp(-a*HP[t,2]))
c(H,P)}
matplot(HP,type="l",lty=1:2)
legend(10,2000,c("H","P"),lty=1:2)
HP<-data.frame(H<-numeric(time),P<-numeric(time))
a=0.1;L=3;K=800;c=3
HP[1,]<-c(50,1)
for (t in 1:20) HP[t+1,]<-{
H<-L*HP[t,1]*exp((K-HP[t,1])/K-a*HP[t,2])
P<-c*HP[t,1]*(1-exp(-a*HP[t,2]))
c(H,P)}
matplot(HP,type="l",lty=1:2)
legend(10,2000,c("H","P"),lty=1:2)
HP<-data.frame(H<-numeric(time),P<-numeric(time))
a=0.005;L=3;K=500;c=2
HP[1,]<-c(50,1)
for (t in 1:20) HP[t+1,]<-{
H<-L*HP[t,1]*exp((K-HP[t,1])/K-a*HP[t,2])
P<-c*HP[t,1]*(1-exp(-a*HP[t,2]))
c(H,P)}
matplot(HP,type="l",lty=1:2)
legend(10,4000,c("H","P"),lty=1:2)