MATH35032: Mathematical Biology, Problem Set 1

This weeks problem set is also available at http://bit.ly/MancMathBio.

Problems (1), (2) and (4) were inspired by originals in Martin Brauns Differential
Equations and Their Applications, Springer Texts in Applied Mathematics, Vol. 11,
Springer-Verlag, (1993), while the last problem, (6), comes from Clifford H. Taubess
Modelling Differential Equations in Biology, 2nd edition, CUP, (2008).
Malthusian Growth
(1) In 1879 and 1881 a number of yearling bass were seined in New Jersey, taken
across North America by train, and planted in San Francisco Bay: a scant 435 young
fish survived the rigours of these journeys. Yet in 1899 the commercial catch of bass
in the Bay area weighed more than 560,000 kilograms in total. As the population
clearly grew very rapidly, it seems reasonable to assume it obeyed the Malthusian
law dN/dt = rN. Assuming the average weight of a bass is 1.5 kg and that 1899s
catch represented 10% of the total population, find a lower bound on r.
(2) Assume that N(t) obeys Malthusian growth. Prove that the increases in N
during successive time intervals of equal duration form the terms of a geometric
progression. This observation is the basis for Thomas Malthuss famous dictum:
Population when unchecked increases in a geometrical ratio. Subsistence
increases only in an arithmetic ratio. A slight acquaintance with numbers
will show the immensity of the first power in comparison with the second.
Logistic Growth
(3) If a population obeys the logistic growth law


dN
N
= rN 1
dt
K

(3.1)

and has initial population N(t = 0) = N0 verify the solution

N(t) =

N0 ert
(1 (N0 /K)(1 ert ))

(4) Mice in the wild obey the logistic population law (3.1). Plagues arise in mouse
populations whenever N(t) becomes too large and an epidemic can destroy 97-98%
of the population within just a few weeks. After such a disaster the mice, now
diminished to perhaps 2% of their pre-epidemic numbers, become too isolated to
permit the disease to spread. The population therefore begins to grow again until
it reaches a level susceptible to another plague.
One can model this situation as follows: say that an epidemic starts when N(t)
reaches some level Q with 0 Q < K. One then models the plague by drastically

reducing the carrying capacity in (3.1): K K with 1 and K Q so the

equation governing a population collapse is:


N
dN
N
(4.1)
=r 1
dt
K
Once the population has declined far enoughsay to a level q with K < q Q
the epidemic burns out and the population begins to recover, once again evolving
according to (3.1).
(a) Mice reproduce very rapidly, so during the recovery phase after an epidemic
it is a good approximation to neglect the quadratic term in (3.1). Solve the
resulting equation and obtain an expression for the time T1 that it takes for
the mouse population to grow from q to Q.
(b) Now consider the epidemic phase, (4.1). Imagine that the plague is so devastating that one can neglect the linear (birth) term and solve the resulting
equation for N(t) with N(0) = Q. Find an expression for the time T2 that it
takes for the population to fall to q.
(c) Sketch a graph showing the temporal evolution of a mouse population that
starts with 0 N(0) < q. Your sketch should include a few cycles of epidemics
followed by recoveries.
(d) Assuming that the time taken to recover from an epidemic is about 4 years
and that Q/q is about 50, show that the intrinsic growth rate r is about 1
mouse per year. This value of r, incidentally, corresponds well with the rate
of reproduction of mice observed in natural circumstances.
Delay Differential Equations
This problem comes from J. D. Murrays Mathematical Biology I: An Introduction.
I have changed the notation slightly so that the analysis resembles more closely the
one we did in lecture.
(5) A continuous time model for the population of baleen whales (a more complicated version of which has been used by the International Whaling Commission) is
the differential delay equation
z 



dN
N(t T )
.
(5.1)
= N(t) + N(t T ) 1 + q 1
dt
K
Here > 0 is a measure of mortality, q > 0 is the a measure of the possible increase
in fecundity (offspring born per parent whale), K > 0 is, as usually, the carrying
capacity of the environment, the delay T is the time it takes a whale to reach sexual
maturity and z > 0 controls the abruptness with which the birth rate responds to
changes in density.

(a) Explore the role of the parameter z by drawing curves of the form
 z
N
f (N) = 1
K
where, on each curve, z is held constant and 0 N 1.1 K. Draw a
separate curve for each value of: z {0.1, 1, 10}.
(b) Determine the steady-state solutions (those with N(t) = a constant) of (5.1).
(c) Show that the equation governing small perturbations n(t) near the positive
steady-state population obey the (linearized) equation
dn(t)
n(t) (qz 1)n(t T ).
dt

(5.2)

(d) Seek solutions to (5.2) of the form n(t) = n0 exp(t) and show that any such
solution must have
= (qz 1)eT .
(5.3)
Considering C and defining = + i, find the real and imaginary parts
of (5.3).
(e) Find conditions under which it is possible that there is an exponentially growing solution to (5.3). That is, find conditions on the parameters that permit
solutions with > 0. In particular, by solving for the limiting case = 0,
show that if (qz 1) > 1, there is a critical delay Tc such that if T < Tc ,
then < 0. Show also that if (qz 1) < 1, then < 0 for all values of the
delay T .
(f) Now concentrate on the potentially unstable regime, where (qz 1) > 1, and
define
b = (qz 1).
Consider delays just slightly greater than Tc . That is, put T = Tc + , where
0 < 1, into (5.3) and, by finding , show that the period of the unstable
oscillation is
2
2
=
+ O().

b2 1
Symmetry and Snail Shells
This last problem was inspired by a set of three papers reprinted in Taubess book.
The first, (Gould 1995), is by the great naturalist Stephen J. Gould and points
out that the vast majority of know species of snails have shells that are coiled in a
right-handed sense, where right-handed is defined as described in Figure 1 below.
The other two articles (Mochizuki et al. 1998, Vogan & Tabin 1999) have to
do with the biological basis of this asymmetry: the biological reality is a lot more
complicated and subtle than models such as the one sketched below might suggest.

Figure 1: A shell is said to be right-handed, or dextral if it is coiled as the one on

the right above, and left-handed or sinistral otherwise.
(6) Taubes proposes a model for the evolution of handedness in a population of
animals where shells of both kinds appear and he bases it on a pair of assumptions:
Snails choose their mates at random. Thus the probability that any given snail
will choose a left-coiling mate is
=

L
,
L+R

where L and R are, respectively, the numbers of left- and right-coiling snails.
The snails mate at a rate , so that the probability that a given snail mates
in a brief interval of length t is t.
On average, each mating gives rise to m new snails.
If two left-coiling snails mate, they produce left-coiling offspring. Similarly, a
mating between two right-coiling snails produces right-coiling offspring. But
a mating involving one snail of each type produces offspring of each type with
equal probability.
The resulting model is

d
= (1 )( 1/2)
dt
where is the fraction of left-coiling snails

(6.1)

(a) Find all the equilibria of (6.1) and classify them according to stability.
(b) Derive (6.1) from the assumptions above.

References
Gould, S. J. (1995), Left snails and right minds, Natural History 104(4), 1018.

Mochizuki, T., Saijoh, Y., Tsuchiya, K., Shirayoshi, Y., Takai, S., Taya, C.,
Yonekawa, H., Yamada, K., Nihei, H., Nakatsuji, N., Overbeek, P., Hamada, H.
& Yokoyama, T. (1998), Cloning of inv, a gene that controls left/right asymmetry and kidney development, Nature 395(6698), 177181.
URL: http://dx.doi.org/10.1038/26006
Vogan, K. & Tabin, C. (1999), Developmental biology - a new spin on handed
asymmetry, Nature 397(6717), 295.
URL: http://dx.doi.org/10.1038/16796