Calculus3Differential Equations6 of 7
Differential Equations · 6

Logistic Growth

1

Growth with a limit

Let's find a differential equation for a population that grows like an exponential at first, but approaches the most a lake can hold. Say our lake holds at most 1000\cc{1000} fish, and we start with 100100 fish at time t=0t = 0, with tt in months. With pure exponential growth, the population passes 1000\cc{1000} and keeps growing forever.

dPdt\displaystyle \rate{\frac{dP}{dt}} =0.5P,P(0)\displaystyle = 0.5P, \quad P(0) =100\displaystyle = 100
→  P\displaystyle \to\; P =100e0.5t\displaystyle = 100e^{0.5t}
100e0.5t\displaystyle 100e^{0.5t} =1000\displaystyle = \cc{1000}
→  t\displaystyle \to\; t =2ln⁡10\displaystyle = 2\ln 10

We want a rate that's close to 0.5P0.5P while there's plenty of room, and 0\rate{0} once the lake is full. So we multiply 0.5P0.5P by the fraction of the lake that's still free.

Unlimited growth
dPdt\displaystyle \rate{\frac{dP}{dt}} =0.5P\displaystyle = 0.5P
Growth limited by room
dPdt\displaystyle \rate{\frac{dP}{dt}} =0.5P(1−P1000)\displaystyle = 0.5P\room{\left(1 - \frac{P}{1000}\right)}
Find a differential equation for a population limited to 1000 fish
dPdt=  ?\displaystyle \rate{\frac{dP}{dt}} = \;?
tt
PP
100100
1000\htmlClass{lc-blue}{1000}
Text solution
The fraction of the lake still freeWith PP fish, there's room for 1000−P\room{1000 - P} more. Dividing by 10001000 gives the fraction that's free.
1000−P1000\displaystyle \room{\frac{1000 - P}{1000}} =1−P1000\displaystyle = \room{1 - \frac{P}{1000}}
Check the endsWhen PP is small, the fraction is close to 11, so the rate is close to 0.5P0.5P. At P=1000P = \cc{1000} the fraction is 0\room{0}, so the rate is 0\rate{0}.
0.5⋅100⋅0.9\displaystyle 0.5\cdot100\cdot\room{0.9} =45\displaystyle = \rate{45}
0.5⋅1000⋅0\displaystyle 0.5\cdot1000\cdot\room{0} =0\displaystyle = \rate{0}

Here's the new rate at a few populations. It starts close to the exponential rate 0.5P0.5P, falls back to 0\rate{0} at 1000\cc{1000}, and is negative when there are too many fish.

PP0.5P0.5P1−P1000\room{1 - \dfrac{P}{1000}}dPdt\rate{\dfrac{dP}{dt}}
10010050500.9\room{0.9}45\rate{45}
5005002502500.5\room{0.5}125\rate{125}
9009004504500.1\room{0.1}45\rate{45}
100010005005000\room{0}0\rate{0}
12001200600600−0.2\room{-0.2}−120\rate{-120}

Because dPdt=0\rate{\frac{dP}{dt}} = \rate{0} at P=1000P = \cc{1000}, we call 1000\cc{1000} the carrying capacity, written as M\cc{M}. Replacing 1000\cc{1000} with M\cc{M} and 0.50.5 with kk gives the logistic equation. When PP is near 00, it's approximately kPkP, so the population grows like an exponential at the start.

The logistic equation
dPdt\displaystyle \rate{\frac{dP}{dt}} =kP(1−PM)\displaystyle = kP\room{\left(1 - \frac{P}{M}\right)}

You'll also see it with 11000\frac{1}{1000} factored out. Then the rate is proportional to both the population PP and the room left, M−P\room{M - P}.

0.5P(1−P1000)\displaystyle 0.5P\room{\left(1 - \frac{P}{1000}\right)} =0.0005P(1000−P)\displaystyle = 0.0005P\room{(1000 - P)} =0.5P−0.0005P2\displaystyle = 0.5P - 0.0005P^2
Rate proportional to population and room left
dPdt\displaystyle \rate{\frac{dP}{dt}} =kP(M−P)\displaystyle = kP\room{(M - P)}

In any form, the carrying capacity M\cc{M} is the non-zero population where dPdt=0\rate{\frac{dP}{dt}} = \rate{0}. In the second form, don't reach for 10.002=500\frac{1}{0.002} = 500. The rate is 00 when 600−P=0\room{600 - P} = 0, so M=600\cc{M} = \cc{600}.

0.3P(1−P800)\displaystyle 0.3P\room{\left(1 - \frac{P}{800}\right)} =0\displaystyle = 0
→  M\displaystyle \to\; \cc{M} =800\displaystyle = \cc{800}
0.002P(600−P)\displaystyle 0.002P\room{(600 - P)} =0\displaystyle = 0
→  M\displaystyle \to\; \cc{M} =600\displaystyle = \cc{600}
3P−0.01P2\displaystyle 3P - 0.01P^2 =P(3−0.01P)\displaystyle = P\room{(3 - 0.01P)} =0\displaystyle = 0
→  M\displaystyle \to\; \cc{M} =300\displaystyle = \cc{300}
2

Where the population goes, and where it grows fastest

Where does the population end up if we start at 100100 fish, 12001200 fish, 00 fish, or 10001000 fish? And at what population are the fish growing fastest? We can answer both questions from the differential equation, without solving it.

Notice there's no tt on the right side of dPdt=0.5P(1−P1000)\rate{\frac{dP}{dt}} = 0.5P\room{\left(1 - \frac{P}{1000}\right)}, so the slope depends only on PP. That means one graph of dPdt\rate{\frac{dP}{dt}} against PP gives the slope at every tt.

PP
dP/dt\htmlClass{lc-green}{dP/dt}
P=1000P = \htmlClass{lc-blue}{1000}
00
100100
500500
900900
12001200
125\htmlClass{lc-green}{125}
45\htmlClass{lc-green}{45}
45\htmlClass{lc-green}{45}
−120\htmlClass{lc-green}{-120}

If we turn the rate graph a quarter turn so PP is vertical, a horizontal line at any height PP meets both graphs, and the rate graph gives the slope of P(t)P(t) at that height.

tt
dP/dt\htmlClass{lc-green}{dP/dt}
PP
PP
100100
500500
1000\htmlClass{lc-blue}{1000}
125\htmlClass{lc-green}{125}
Find the long-term population and fastest growth from the rate graph
dPdt=0.5P(1−P1000)\displaystyle \rate{\frac{dP}{dt}} = 0.5P\room{\left(1 - \frac{P}{1000}\right)}
tt
PP
00
100100
12001200
1000\htmlClass{lc-blue}{1000}
Text solution
Start at 100100The rate is positive, so PP increases. It speeds up until the vertex, then slows as dPdt\rate{\frac{dP}{dt}} falls toward 0\rate{0}, so PP approaches 1000\cc{1000}.
P\displaystyle P =100 ⁣:dPdt\displaystyle = 100\colon\quad \rate{\frac{dP}{dt}} =0.5⋅100⋅0.9\displaystyle = 0.5\cdot100\cdot\room{0.9} =45>0\displaystyle = \rate{45} \mathrel{\rate{>}} \rate{0}
Start at 12001200, 00 or 10001000Above 1000\cc{1000} the rate is negative, so PP decreases toward 1000\cc{1000}. At 00 and 1000\cc{1000} the rate is 0\rate{0}, so PP never changes.
P\displaystyle P =1200 ⁣:dPdt\displaystyle = 1200\colon\quad \rate{\frac{dP}{dt}} =0.5⋅1200⋅(−0.2)\displaystyle = 0.5\cdot1200\cdot\room{(-0.2)} =−120\displaystyle = \rate{-120}
Fastest at the vertexAt 100100 and 900900 we multiply the same two numbers, 100100 and 900\room{900}, so the rates match. The vertex is halfway between the zeros 00 and 1000\cc{1000}.
0.0005⋅100⋅900\displaystyle 0.0005\cdot100\cdot\room{900} =45\displaystyle = \rate{45}
0.0005⋅900⋅100\displaystyle 0.0005\cdot900\cdot\room{100} =45\displaystyle = \rate{45}
P\displaystyle P =500 ⁣:dPdt\displaystyle = 500\colon\quad \rate{\frac{dP}{dt}} =0.5⋅500⋅0.5\displaystyle = 0.5\cdot500\cdot\room{0.5} =125\displaystyle = \rate{125}

So the sign of the rate tells us where each start goes.

0<P<1000:\displaystyle 0 < P < \cc{1000}:
dPdt>0  →  P increasing\displaystyle {\rate{\frac{dP}{dt} > 0} \;\to\; P \text{ increasing}}
P>1000:\displaystyle P > \cc{1000}:
dPdt<0  →  P decreasing\displaystyle {\rate{\frac{dP}{dt} < 0} \;\to\; P \text{ decreasing}}
P\displaystyle P =0 or P\displaystyle = 0 \text{ or } P =1000:\displaystyle = \cc{1000}:
dPdt\displaystyle \rate{\frac{dP}{dt}} =0  →  P constant\displaystyle = \rate{0} \;\to\; P \text{ constant}
For any positive start
P(0)>0  →  lim⁡t→∞P(t)\displaystyle P(0) > 0 \;\to\; \lim_{t\to\infty} P(t) =1000\displaystyle = \cc{1000}

And growth is fastest at half the carrying capacity. That's a population, not a time. The horizontal line at P=500P = 500 passes through the vertex of the parabola and the steepest point of the solution curve.

fastest growth at P\displaystyle \text{fastest growth at } P =M2\displaystyle = \frac{\cc{M}}{2}
3

Concavity and inflection points

We know any positive start approaches 1000\cc{1000}, and the population grows fastest at P=500P = 500. Next, where is the curve concave up, and where is it concave down? That comes from the sign of d2Pdt2\frac{d^2P}{dt^2}, and since PP is a function of tt, finding it takes the chain rule.

Find where solutions to dPdt=0.5P(1−P1000)\rate{\frac{dP}{dt}} = 0.5P\room{\left(1 - \frac{P}{1000}\right)} are concave up and concave down
tt
dP/dt\htmlClass{lc-green}{dP/dt}
PP
PP
100100
500500
1000\htmlClass{lc-blue}{1000}
125\htmlClass{lc-green}{125}
Text solution
Differentiate with the chain ruleWrite the rate as 0.5P−0.0005P20.5P - 0.0005P^2, differentiate with respect to PP, then multiply by dPdt\rate{\frac{dP}{dt}}.
d2Pdt2\displaystyle \frac{d^2P}{dt^2} =(0.5−0.001P)⋅dPdt\displaystyle = (0.5 - 0.001P)\cdot\rate{\frac{dP}{dt}}
The sign of each factorThe first factor is the slope of the rate parabola: positive below the vertex P=500P = 500, negative above it. The second is the rate itself: positive between 00 and 1000\cc{1000}, negative above 1000\cc{1000}.
Above 1000\cc{1000}, don't stop at the first factorAt P=1200P = 1200 the first factor is negative, but so is the rate, so the curve is concave up.
d2Pdt2\displaystyle \frac{d^2P}{dt^2} =(−0.7)(−120)\displaystyle = (-0.7)\rate{(-120)} =84>0\displaystyle = \cvup{84} \mathrel{\cvup{>}} \cvup{0}

Multiplying the signs of the two factors gives the concavity in each band of PP.

0<P<500:\displaystyle 0 < P < 500:
(+)(+)>0  →  concave up\displaystyle (+)\rate{(+)} \mathrel{\cvup{>}} \cvup{0} \;\to\; \cvup{\text{concave up}}
500<P<1000:\displaystyle 500 < P < \cc{1000}:
(−)(+)<0  →  concave down\displaystyle (-)\rate{(+)} \mathrel{\cvdn{<}} \cvdn{0} \;\to\; \cvdn{\text{concave down}}
P>1000:\displaystyle P > \cc{1000}:
(−)(−)>0  →  concave up\displaystyle (-)\rate{(-)} \mathrel{\cvup{>}} \cvup{0} \;\to\; \cvup{\text{concave up}}

A solution never crosses 1000\cc{1000}, so the only place its concavity can change is at P=500P = 500. A curve has an inflection point only if its population passes through 500500.

tt
PP
100100
500500
700700
12001200
1000\htmlClass{lc-blue}{1000}
concave up\htmlClass{lc-orange}{\text{concave up}}
concave down\htmlClass{lc-violet}{\text{concave down}}
concave up\htmlClass{lc-orange}{\text{concave up}}
P(0)\displaystyle P(0) =100:\displaystyle = 100:
concave up to P\displaystyle \cvup{\text{concave up}} \text{ to } P =500, concave down after\displaystyle = 500,\ \cvdn{\text{concave down}} \text{ after}
→  inflection point at P\displaystyle \to\; \text{inflection point at } P =500\displaystyle = 500
concave up to P\displaystyle \cvup{\text{concave up}} \text{ to } P =500,\displaystyle = 500,
concave down after\displaystyle \cvdn{\text{concave down}} \text{ after}
→  inflection point at P\displaystyle \to\; \text{inflection point at } P =500\displaystyle = 500
P(0)\displaystyle P(0) =700:\displaystyle = 700:
500<P<1000  →  concave down,\displaystyle 500 < P < \cc{1000} \;\to\; \cvdn{\text{concave down}},
no inflection point\displaystyle \text{no inflection point}
P(0)\displaystyle P(0) =1200:\displaystyle = 1200:
P>1000  →  concave up\displaystyle P > \cc{1000} \;\to\; \cvup{\text{concave up}}

How does the initial population change growth and concavity?

4

Finding the formula for P(t)P(t)

We know the population starts at 100100, approaches 1000\cc{1000}, and increases fastest at 500500. Now let's find the exact formula for P(t)P(t). The equation is separable, and we'll need partial fractions on the way.

Solve dPdt=0.5P(1−P1000)\rate{\frac{dP}{dt}} = 0.5P\room{\left(1 - \frac{P}{1000}\right)} with P(0)=100P(0) = 100
dPdt\displaystyle \rate{\frac{dP}{dt}} =0.5P(1−P1000)\displaystyle = 0.5P\room{\left(1 - \frac{P}{1000}\right)}
P(0)\displaystyle P(0) =100\displaystyle = 100
P(t)\displaystyle P(t) =  ?\displaystyle = \;?
Text solution
Separate, then split by partial fractionsFactor out 11000\frac{1}{1000} to clear the fraction, then divide by P(1000−P)P\room{(1000 - P)}. Adding 1P\frac{1}{P} and 11000−P\room{\frac{1}{1000 - P}} gives a numerator of 10001000, so we divide by 10001000.
dPP(1000−P)\displaystyle \frac{dP}{P\room{(1000 - P)}} =0.0005 dt\displaystyle = 0.0005\,dt
1P(1000−P)\displaystyle \frac{1}{P\room{(1000 - P)}} =11000(1P+11000−P)\displaystyle = \frac{1}{1000}\left(\frac{1}{P} + \room{\frac{1}{1000 - P}}\right)
Integrate, minding the signThe derivative of 1000−P\room{1000 - P} is −1-1, so its integral brings a minus. Both PP and 1000−P\room{1000 - P} stay positive, so no absolute values.
ln⁡P1000−P\displaystyle \ln\frac{P}{\room{1000 - P}} =0.5t+1000C\displaystyle = 0.5t + 1000C
P1000−P\displaystyle \frac{P}{\room{1000 - P}} =Be0.5t\displaystyle = Be^{0.5t}
Flip it, so PP appears onceThe reciprocal splits into 1000P−1\frac{\cc{1000}}{P} - 1. At t=0t = 0, it's the room left over the fish at the start, 900100=9\frac{\room{900}}{100} = 9.
1000P−1\displaystyle \frac{\cc{1000}}{P} - 1 =9e−0.5t\displaystyle = 9e^{-0.5t}
P\displaystyle P =10001+9e−0.5t\displaystyle = \frac{\cc{1000}}{1 + 9e^{-0.5t}}
Check it, and find the fastest growthAt t=0t = 0 it gives 100010=100\frac{\cc{1000}}{10} = 100, and as t→∞t \to \infty it approaches 1000\cc{1000}. It reaches 500500 when the denominator is 22.
9e−0.5t\displaystyle 9e^{-0.5t} =1  →  e0.5t\displaystyle = 1 \;\to\; e^{0.5t} =9  →  t\displaystyle = 9 \;\to\; t =2ln⁡9\displaystyle = 2\ln 9

In the general formula, M\cc{M} is the carrying capacity, AA is the room left over the starting population, and kk is the growth rate while the population is small. When the equation is written as 0.0005P(1000−P)0.0005P\room{(1000 - P)}, the exponent is −0.0005⋅1000 t=−0.5t-0.0005 \cdot \cc{1000}\,t = -0.5t, not −0.0005t-0.0005t.

P(t)\displaystyle P(t) =M1+Ae−kt,A\displaystyle = \frac{\cc{M}}{1 + Ae^{-kt}}, \qquad A =M−P0P0\displaystyle = \frac{\room{M - P_0}}{P_0}
dPdt\displaystyle \rate{\frac{dP}{dt}} =0.0005P(1000−P)\displaystyle = 0.0005P\room{(1000 - P)}
→  e−0.0005⋅1000 t\displaystyle \to\; e^{-0.0005\cdot\cc{1000}\,t} =e−0.5t\displaystyle = e^{-0.5t}

Summary

The logistic equation and carrying capacity
dPdt\displaystyle \rate{\frac{dP}{dt}} =kP(1−PM)\displaystyle = kP\room{\left(1 - \frac{P}{M}\right)}
dPdt\displaystyle \rate{\frac{dP}{dt}} =0.002P(600−P)\displaystyle = 0.002P\room{(600 - P)}
→  M\displaystyle \to\; \cc{M} =600\displaystyle = \cc{600}

The factor 1−PM\room{1 - \frac{P}{M}} is the fraction of capacity still free. When PP is near 00, it is close to 11, so growth is nearly exponential. When P=MP = \cc{M}, no room is left, and dPdt=0\rate{\frac{dP}{dt}} = \rate{0}.

In any form of the equation, the carrying capacity M\cc{M} is the positive population that makes dPdt=0\rate{\frac{dP}{dt}} = \rate{0}. For 0.002P(600−P)=00.002P\room{(600 - P)} = 0, that gives M=600\cc{M} = \cc{600}.

Long-term population
P(0)>0  →  lim⁡t→∞P(t)\displaystyle P(0) > 0 \;\to\; \lim_{t\to\infty} P(t) =M\displaystyle = \cc{M}

For any positive starting population, P(t)P(t) approaches the carrying capacity M\cc{M} as tt goes to infinity.

When 0<P<M0 < P < \cc{M}, the rate dPdt\rate{\frac{dP}{dt}} is positive, so the population increases toward M\cc{M}. When P>MP > \cc{M}, dPdt\rate{\frac{dP}{dt}} is negative, so the population decreases toward M\cc{M}. The values P=0P = 0 and P=MP = \cc{M} give dPdt=0\rate{\frac{dP}{dt}} = \rate{0}, so they are constant solutions.

Fastest growth
fastest growth at P\displaystyle \text{fastest growth at } P =M2\displaystyle = \frac{\cc{M}}{2}

The graph of dPdt\rate{\frac{dP}{dt}} against PP is a downward parabola with zeros at P=0P = 0 and P=MP = \cc{M}. The vertex is halfway between those zeros, so the growth rate is largest at P=M2P = \frac{\cc{M}}{2}. This is a population, not a time. To find the time of fastest growth, solve P(t)=M2P(t) = \frac{\cc{M}}{2} for tt.

Concavity and inflection points
d2Pdt2\displaystyle \frac{d^2P}{dt^2} =(0.5−0.001P)⋅dPdt\displaystyle = (0.5 - 0.001P)\cdot\rate{\frac{dP}{dt}}
0<P<500\displaystyle 0 < P < 500
concave up\displaystyle \cvup{\text{concave up}}
500<P<1000\displaystyle 500 < P < \cc{1000}
concave down\displaystyle \cvdn{\text{concave down}}
P>1000\displaystyle P > \cc{1000}
concave up\displaystyle \cvup{\text{concave up}}

Because PP depends on tt, finding d2Pdt2\frac{d^2P}{dt^2} needs the chain rule. Differentiate dPdt\rate{\frac{dP}{dt}} with respect to PP, then multiply by dPdt\rate{\frac{dP}{dt}}. The sign of d2Pdt2\frac{d^2P}{dt^2} comes from multiplying the signs of these two factors.

For our lake with M=1000\cc{M} = \cc{1000}, solutions are concave up for 0<P<5000 < P < 500, concave down for 500<P<1000500 < P < \cc{1000}, and concave up for P>1000P > \cc{1000}. An inflection point occurs where the curve intersects P=M2P = \frac{\cc{M}}{2}, which happens only when the starting population is between 00 and M2\frac{\cc{M}}{2}.

The solution formula
P(t)\displaystyle P(t) =M1+Ae−kt,A\displaystyle = \frac{\cc{M}}{1 + Ae^{-kt}}, \qquad A =M−P0P0\displaystyle = \frac{\room{M - P_0}}{P_0}

The numerator M\cc{M} is the carrying capacity, and A=M−P0P0A = \frac{\room{M - P_0}}{P_0} is the ratio of room left to starting population at t=0t = 0. Growth is fastest when Ae−kt=1Ae^{-kt} = 1, which makes the denominator 22 and gives P=M2P = \frac{\cc{M}}{2}.

If the differential equation is written as dPdt=kP(M−P)\rate{\frac{dP}{dt}} = kP\room{(M - P)}, the exponent is −kMt-k\cc{M}t.

5

Practice

Give exact values, keeping ln⁡\ln, unless the question asks for an estimate.

1
A solution of dydt=0.3y−0.002y2\dfrac{dy}{dt} = 0.3y - 0.002y^2 has y(0)=20y(0) = 20. (a) Find lim⁡t→∞y(t)\displaystyle\lim_{t\to\infty} y(t). (b) For which starting values y(0)y(0) does the solution curve have an inflection point?

To find the carrying capacity, set 0.3y−0.002y2=00.3y - 0.002y^2 = 0 for y≠0y \ne 0. What is yy?

If you got ∞\infty Check the −0.002y2-0.002y^2 term. It keeps y(t)y(t) from increasing without bound and makes it approach the carrying capacity.

An inflection point happens at y=M2y = \frac{M}{2}. Which starting values make y(t)y(t) pass through M2\frac{M}{2}?

The carrying capacity and limitFactor out 0.002y0.002y to find the carrying capacity M\cc{M}.
0.3y−0.002y2\displaystyle 0.3y - 0.002y^2 =0.002y(150−y)  →  M\displaystyle = 0.002y\room{(150 - y)} \;\to\; \cc{M} =150\displaystyle = \cc{150}
The starting value 2020 is between 00 and 150\cc{150}, so dydt>0\rate{\frac{dy}{dt} > 0} and yy increases toward the carrying capacity 150\cc{150}.
0<20<150:dydt>0  →  lim⁡t→∞y(t)\displaystyle 0 < 20 < \cc{150}: \quad \rate{\frac{dy}{dt} > 0} \;\to\; \lim_{t\to\infty} y(t) =150\displaystyle = \cc{150}
The inflection pointThe curve changes concavity at y=M2=75y = \frac{\cc{M}}{2} = 75, and y(t)y(t) only passes through 7575 if it starts below 7575.
M2\displaystyle \frac{\cc{M}}{2} =75:0<y(0)<75  →  inflection point at y\displaystyle = 75: \quad 0 < y(0) < 75 \;\to\; \text{inflection point at } y =75\displaystyle = 75
2
Let y(t)y(t) be the solution of dydt=y4(10−y)\dfrac{dy}{dt} = \dfrac{y}{4}(10 - y) with y(0)=14y(0) = 14. (a) Find d2ydt2\dfrac{d^2y}{dt^2} in terms of yy. (b) Is the solution curve concave up or concave down? (c) Find lim⁡t→∞y(t)\displaystyle\lim_{t\to\infty} y(t).

Because yy depends on tt, what factor does the chain rule give when you differentiate with respect to tt?

If you got d2ydt2=10−2y4\dfrac{d^2y}{dt^2} = \dfrac{10 - 2y}{4} Remember that yy depends on tt, so the chain rule gives an extra factor of dydt\frac{dy}{dt}.

(b)\text{(b)}

At y=14y = 14, what are the signs of 10−2y4\frac{10 - 2y}{4} and dydt\frac{dy}{dt}?

If you got concave down Check the sign of dydt\frac{dy}{dt} at y=14y = 14. Both factors in d2ydt2\frac{d^2y}{dt^2} are negative, so their product is positive.

The second derivativeMultiply out the numerator, then differentiate with respect to tt using the chain rule.
dydt\displaystyle \rate{\frac{dy}{dt}} =10y−y24\displaystyle = \frac{10y - y^2}{4}
d2ydt2\displaystyle \frac{d^2y}{dt^2} =10−2y4⋅dydt\displaystyle = \frac{10 - 2y}{4}\cdot\rate{\frac{dy}{dt}} =10−2y4⋅y(10−y)4\displaystyle = \frac{10 - 2y}{4}\cdot\rate{\frac{y(10 - y)}{4}}
The limitBecause the starting value 1414 is above the carrying capacity 10\cc{10}, dydt\rate{\frac{dy}{dt}} is negative, so y(t)y(t) decreases toward 10\cc{10}.
y(0)\displaystyle y(0) =14>10:dydt<0  →  y decreasing,lim⁡t→∞y(t)\displaystyle = 14 > \cc{10}: \quad \rate{\frac{dy}{dt} < 0} \;\to\; y \text{ decreasing}, \quad \lim_{t\to\infty} y(t) =10\displaystyle = \cc{10}
The concavityFor y>10y > \cc{10}, both factors in d2ydt2\frac{d^2y}{dt^2} are negative, so their product is positive and the curve is concave up.
y>10:10−2y4<0,y(10−y)4<0  →  d2ydt2>0\displaystyle y > \cc{10}: \quad \frac{10 - 2y}{4} < 0, \quad \rate{\frac{y(10 - y)}{4} < 0} \;\to\; \frac{d^2y}{dt^2} \mathrel{\cvup{>}} \cvup{0}
y\displaystyle y =14:(−4.5)(−14)\displaystyle = 14: \quad (-4.5)\rate{(-14)} =63>0\displaystyle = \cvup{63} \mathrel{\cvup{>}} \cvup{0}
3
Solve dPdt=0.0004P(500−P)\dfrac{dP}{dt} = 0.0004P(500 - P), P(0)=100P(0) = 100.

Separate the variables, then split 1P(500−P)\frac{1}{P(500 - P)} into partial fractions with denominators PP and 500−P500 - P.

Once you have P500−P=Be0.2t\frac{P}{500 - P} = Be^{0.2t}, take the reciprocal of both sides to make solving for PP simpler.

If you got P=5001+4e−0.0004tP = \dfrac{500}{1 + 4e^{-0.0004t}} In dPdt=kP(M−P)\frac{dP}{dt} = kP(M - P), the exponent is −kMt-kMt, not −kt-kt. Here kM=0.0004⋅500=0.2kM = 0.0004 \cdot 500 = 0.2.

If you got P=5001+14e−0.2tP = \dfrac{500}{1 + \frac14 e^{-0.2t}} Check AA at t=0t = 0. A=M−P0P0=500−100100=4A = \frac{M - P_0}{P_0} = \frac{500 - 100}{100} = 4, not 100400\frac{100}{400}.

If you got P=5001−4e−0.2tP = \dfrac{500}{1 - 4e^{-0.2t}} Check the split of 500−PP\frac{500 - P}{P}. Dividing each term by PP gives 500P−1\frac{500}{P} - 1, not 1−500P1 - \frac{500}{P}.

Separate and splitSeparate the variables, then use partial fractions on the left side.
dPP(500−P)\displaystyle \frac{dP}{P\room{(500 - P)}} =0.0004 dt,1P(500−P)\displaystyle = 0.0004\,dt, \qquad \frac{1}{P\room{(500 - P)}} =1500(1P+1500−P)\displaystyle = \frac{1}{500}\left(\frac{1}{P} + \room{\frac{1}{500 - P}}\right)
IntegrateIntegrate both sides, remembering that 1500−P\room{\frac{1}{500 - P}} integrates to −ln⁡(500−P)-\room{\ln(500 - P)}, then multiply by 500500.
1500(ln⁡P−ln⁡(500−P))\displaystyle \frac{1}{500}\big(\ln P - \room{\ln(500 - P)}\big) =0.0004t+C  →  ln⁡P500−P\displaystyle = 0.0004t + C \;\to\; \ln\frac{P}{\room{500 - P}} =0.2t+500C\displaystyle = 0.2t + 500C
Take the reciprocal and find the constantTake the exponential of both sides, take the reciprocal of both sides, and plug in t=0t = 0 and P=100P = 100 to find AA.
P500−P\displaystyle \frac{P}{\room{500 - P}} =Be0.2t  →  500−PP\displaystyle = Be^{0.2t} \;\to\; \frac{\room{500 - P}}{P} =Ae−0.2t,A\displaystyle = Ae^{-0.2t}, \qquad A =1B\displaystyle = \frac{1}{B}
t\displaystyle t =0:500−100100\displaystyle = 0: \quad \frac{\room{500 - 100}}{100} =4\displaystyle = 4 =A\displaystyle = A
Solve for PPSplit 500−PP\frac{\room{500 - P}}{P} into 500P−1\frac{\cc{500}}{P} - 1, add 11, and solve for PP.
500P−1\displaystyle \frac{\cc{500}}{P} - 1 =4e−0.2t  →  P\displaystyle = 4e^{-0.2t} \;\to\; P =5001+4e−0.2t\displaystyle = \frac{\cc{500}}{1 + 4e^{-0.2t}}
4
A population is modelled by P(t)=20001+19e−0.4tP(t) = \dfrac{2000}{1 + 19e^{-0.4t}}. Find P(0)P(0), the carrying capacity, and the time at which the population is growing fastest.

The fastest growth occurs when the population reaches half the carrying capacity, M2=1000\frac{M}{2} = 1000. What is tt then?

If you got 10001000 for the time 10001000 is the population where the growth is fastest, but the problem asks for the time tt when that happens.

Plug in t=0t = 0 to get P(0)P(0), and take the limit as t→∞t \to \infty to find the carrying capacity M\cc{M}.
P(0)\displaystyle P(0) =20001+19\displaystyle = \frac{\cc{2000}}{1 + 19} =100,M\displaystyle = 100, \qquad \cc{M} =2000\displaystyle = \cc{2000}
The time of fastest growthThe population grows fastest when it reaches half the carrying capacity, so set P(t)=1000P(t) = 1000 and solve for tt.
20001+19e−0.4t\displaystyle \frac{\cc{2000}}{1 + 19e^{-0.4t}} =1000  →  19e−0.4t\displaystyle = 1000 \;\to\; 19e^{-0.4t} =1  →  t\displaystyle = 1 \;\to\; t =ln⁡190.4\displaystyle = \frac{\ln 19}{0.4} =2.5ln⁡19\displaystyle = 2.5\ln 19
5
A population follows P(t)=10001+Ae−ktP(t) = \dfrac{1000}{1 + Ae^{-kt}} with P(0)=200P(0) = 200 and P(2)=500P(2) = 500. Find kk.

Use P(0)=200P(0) = 200 to find A=M−P0P0A = \frac{M - P_0}{P_0}.

Plug in t=2t = 2 and set the denominator equal to 1000500=2\frac{1000}{500} = 2.

Use the initial population P(0)=200P(0) = 200 to find AA.
A\displaystyle A =1000−200200\displaystyle = \frac{\room{1000 - 200}}{200} =4\displaystyle = 4
Plug in t=2t = 2 and P=500P = 500, then isolate e−2ke^{-2k} and take the natural log of both sides.
10001+4e−2k\displaystyle \frac{\cc{1000}}{1 + 4e^{-2k}} =500  →  4e−2k\displaystyle = 500 \;\to\; 4e^{-2k} =1  →  e−2k\displaystyle = 1 \;\to\; e^{-2k} =14  →  k\displaystyle = \frac14 \;\to\; k =ln⁡42\displaystyle = \frac{\ln 4}{2} =ln⁡2\displaystyle = \ln 2
6
A rumour spreads through a school of 1200 students at a rate proportional to the product of the number of students who have heard it and the number who haven't. (a) Write a differential equation for N(t)N(t), the number of students who have heard it. (b) How many students have heard the rumour when it is spreading fastest?

If NN students have heard the rumour, how many students have not heard it yet?

The rate dNdt\frac{dN}{dt} is a downward-opening parabola with zeros at N=0N = 0 and N=1200N = 1200. Where is its vertex?

The number of students who haven't heard the rumour is 1200−N\room{1200 - N}, and the rate is proportional to the product of NN and 1200−N\room{1200 - N}.
dNdt\displaystyle \rate{\frac{dN}{dt}} =kN(1200−N)\displaystyle = kN\room{(1200 - N)}
The rate is a downward parabola with zeros at 00 and 1200\cc{1200}, so its maximum is halfway between them, at N=600N = 600.
dNdt\displaystyle \rate{\frac{dN}{dt}} =0 at N\displaystyle = \rate{0} \text{ at } N =0, N\displaystyle = 0,\ N =1200  →  largest dNdt at N\displaystyle = \cc{1200} \;\to\; \text{largest } \rate{\frac{dN}{dt}} \text{ at } N =600\displaystyle = 600