Professional Documents
Culture Documents
Abundance or Density
Have boundaries
Change over time
Composition (sex and age)
Distribution (pattern, scale)
Population Dynamics
Balance of biology and
mathematics
Population Mathematics
Little or no concern about
biological reality
The bottom line is that models are abstractions of reality. Levins (1966) points out that
modeling is essentially a tradeoff between
1. Generality
2. Realism
3. Precision
Williams et al. (2002, Chapter 7) Describe model usefulness with the following
conceptual model.
A mathematical aside
Change, difference, and ratio equations
When time periods are discrete we use difference equations to describe change over time. Difference
equations can be expressed different ways to highlight various aspects of the math. Three important
ones:
Change Equation
N t +1 = f ( N t )
Difference Equation
N t +1 N t = f ( N t )
Ratio Equation
N t +1
Nt
= f (Nt )
Bt
Nt
and
dt =
Dt
Nt
Substituting
N t +1 = N t + bt N t d t N t
Now we assume that per capita birth and death rates are constant over time
N t +1 = N t + bN t dN t
Simplifying this equation
N t +1 = N t + (b d )N t
The term b-d is the geometric rate of increase and is given its own symbol R
N t +1 = N t + RN t
N t +1 N t = RN t
(a difference equation)
N t = RN t
N t
=R
N
The parameter R represents the per capita rate of change in the size of the population
Lets go back to N t +1 = N t + RN t and factor out Nt
(1 + R ) =
N t +1 = N t
a change equation
N t = N t 1 = 2 N t 2 = 3 N t 3 ... = t N 0
N t = t N 0
lam=0.8
Time
N t +1
Nt
a ratio equation
Hence we can calculate lamda without knowing the per capita birth and death rates;
however, in actuality it will probably be important to estimate b and d.
Let b-d=r
dN
= rN
dt
Put in terms of N
dN
= rdt
N
from calculus
dx
= d ln(x )
x
so
d ln( N ) = rdt
t
d ln( N ) = rdt
Integrate
ln( N t ) ln( N 0 ) = rt
take exponential of both sides
e ln( N ( t )) e ln( N ( 0 )) = e rt
Nt
= e rt
N0
Solve for N(t) and we get a continuous model that predicts population size
N t = N 0 e rt
1
a
r=-0.3
r=0
r=0.3
Time
N t = t N 0
Continuous
N t = N 0 e rt
= e r or ln( ) = r
if r is small then we can use a Taylor series to find the approximate relationship
er = 1 + r +
r2 r3
+ + ...
2! 3!
1+ r
The two models diverge as r gets large and as the number of generations (t) increases.
Consider the following example with R=0.2=r.
0.2
N
t
0
1
2
3
4
5
6
7
8
9
10
= N
10.00
12.00
14.40
17.28
20.74
24.88
29.86
35.83
43.00
51.60
61.92
( )
= N
rt
10.00
12.21
14.92
18.22
22.26
27.18
33.20
40.55
49.53
60.50
73.89
The distinction between R (or ) and r is important biologically, i.e. pulse versus
continuous reproduction, instantaneous versus finite survival rates. There is confusion in
the literature between r, R, and , so be careful to understand the authors definition.