Studying Adaptation
Evolutionary Analysis of Form and
Function
Ch. 9
1
Analysis of Adaptation
• A trait, or integrated suite of traits, that increases the
fitness of its possessor is called an adaptation and is said to
be adaptive
• How do we know that a trait is adaptive?
• The adaptive values of some traits, such as eyes, would not
seem to need much explanation
• The adaptive values of some other traits may be more
subtle — and some traits might not be adaptive at all
• We should be wary of ―adaptive storytelling‖, accepting
plausible hypotheses about adaptation uncritically, and
wary of the ―adaptationist program‖ — the idea that all
traits are somehow adaptive
2
The giraffe’s neck: an adaptive story
• An adaptive story
– Giraffes evolved long necks (and legs) because taller
individuals could browse vegetation that was above the
reach of competitors; or because it allowed them to
exploit a resource that was not available closer to the
ground
• Prediction
– If giraffes are using their height to avoid competition
with other browsers or to exploit a preferred resource,
they should feed ―high‖
3
Do giraffes use their height to forage high? (Fig. 9.2)
4
The giraffe’s neck: observations and an
alternative hypothesis
• Giraffes do most of their browsing on vegetation that is at
shoulder height
• Simmons and Scheepers (1996) suggest that long necks
evolved for use in male – male contests for access to
females
– larger males (especially males with thicker necks and more
massive horns and skulls) are more likely to win contests with
other males
– females are more receptive to courtship by males with the stoutest
necks, and most massive horns and skulls (= more mating success
by larger males?)
– this is an argument that long necks evolved through sexual
selection: both male – male competition and female choice
5
Male giraffes ―neck westling‖
6
Neck size and social interactions in giraffes
(adapted from Table 9.1)
Males in class C are young adults; males in classes A and B are more mature.
Class A males are often larger than class B males, but more importantly, class A
males have stouter necks, more massive horns, and more heavily armored skulls
a. Neck size and male social interactions.
These numbers represent observations of one male displacing another from a social group
A displaces B, A A displaces A, B B displaces A, C
displaces C, or B displaces B, or C displaces A, or C
displaces C displaces C displaces B
82 39 6
b. Neck size and female choice.
These numbers represent male success in obtaining female cooperation in determining
female reproductive state
Successful Unsuccessful % Successful
A bulls 34 22 60.7
B bulls 76 61 55.5
C bulls 45 89 33.6 7
Okapi – the giraffe’s closest living relative
(is this what ancestral giraffes looked like?)
(do Okapi males fight with their necks and heads?)
8
Caveats
• Differences among populations or species are not always
adaptive
– Giraffes from different populations have different spot patterns
— are these differences adaptive or accidental?
• Not every trait, or every use of a trait, is an adaptation
– Giraffes do sometimes forage at their full height, but this is not
adaptive unless it increases fitness (and even if it does, it’s not
necessarily the reason that long necks evolved in the first place)
• Not every adaptation is perfect
– Long necks may help male giraffes get mates, but they make it
hard to take a drink
9
Approaches to studying adaptation
• Experimental manipulation
– what happens when a phenotype is experimentally
altered?
• Observational studies
– do certain phenotypes have higher fitness than others,
or do individuals behave in a way that is consistent with
an adaptive expectation?
• Comparative method
– does the same trait evolve repeatedly in related species
that share similar environments?
10
Experiment
Wing markings and wing-waving display in the tephritid fly
Zonosemata (Green et al. 1987)
• The teprhritid fly Zonosemata vittigera has dark bands on
its wings
• When disturbed, the fly holds out its wings and waves
them up and down
• Seems to mimic the leg waving territorial threat display
of some species of jumping spider
• Hypotheses
1. Not mimicry of jumping spiders
2. Used to deter predators other than jumping spiders
3. Used to deter jumping spiders specifically
11
Experimental treatments for testing the function of
wing-waving display by Zonosemata – 1 (Fig. 9.5)
12
Experimental treatments for testing the function of
wing-waving display by Zonosemata – 2 (Fig. 9.5)
13
Zonosemata mimic jumping spiders to avoid
predation and both components of the mimicry must
be present (Fig. 9.6)
14
Observational studies
Behavioral thermoregulation in iguanas and garter
snakes
• In ectothermic (cold-blooded) organisms, performance is
tied closely to body temperature. In general, there is a
relatively narrow temperature range in which endurance,
speed, digestive efficiency and other physiological and
sensory functions are maximized
• It would seem to be adaptive, therefore, for animals to
regulate their body temperature behaviorally in order to
maintain it in or near the zone that maximizes performance
15
Thermoregulation by desert iguanas (Fig. 9.8)
(Huey and Kinsolver, 1989)
16
Behavioral thermoregulation in garter snakes
(Huey et al., 1989)
• Preferred temperature range in laboratory 28 – 32 ˚C
• Implanted wild snakes with radio transmitters that made it
possible to locate snakes and to record their temperature
• Snakes maintain body temperature within or very close to
their preferred range despite wide fluctuations in
environmental temperature
• Snakes preferentially use micro-environments with
temperatures in their preferred range
17
Body temperatures
of garter snakes in
nature (Fig. 9.9)
(Huey et al., 1989)
18
Environmental
temperatures available
to garter snakes – 1
(Figs. 9.10a-c)
(Huey et al., 1989)
19
Environmental
temperatures available
to garter snakes – 2
(Figs. 9.10d-e)
20
Distributions of rocks available to snakes versus
rocks chosen by snakes (Table 9.2) (Huey et al., 1989)
Thin Medium Thick
( 40 cm)
Rocks
available to 32.4 % 34.6 % 33 %
snakes
Rocks chosen
7.7 % 61.5 % 30.8 %
by snakes
21
The Comparative Method
Testis size and social group size in fruit bats and flying foxes
(Hosken, 1998)
• Relative testis size varies among bat species
• Are large testes an adaptation for sperm competition?
• Sperm competition can occur when a female mates with
two or more males during her fertile period
• Larger testes may mean more sperm, which may increase
chances for successful fertilization under sperm
competition
• Prediction:
– relative testis size will be larger in species that roost in larger
social groups (under the assumption that females in larger groups
are more likely to mate multiple times)
22
Variation in testis size and social group size
among fruit bats and flying foxes (Fig. 9.11b)
23
The comparative method and phylogeny
• The comparative method must take into
account phylogenetic relationships among
the taxa being studied
• This can be done using Felsenstein’s
method of phylogenetically independent
contrasts
24
Pseudoreplication in
comparative analyses
(Fig. 9.12)
If the common ancestor of taxa A-C had
small testis size and small group size, then
these three taxa represent only one
evolutionary acquisition of small testis size
and small group size. Likewise if the
common ancestor of taxa D-F had large
testis size and large group size
25
Felsenstein’s method of evaluating comparative hypotheses
using phylogenetically independent contrasts (Fig. 9.13)
26
Phylogeny, group size, body mass, and testis mass in
fruit bats and flying foxes (Fig. 9.14a)
27
Correlated evolution of testis size and group size in
fruit bats and flying foxes (Fig. 9.14b, c)
28
Testis size and social group size in bats:
Is it really sperm competition?
• What additional data would help to test the
hypothesis that sperm competition drives
the observed positive relationship between
social group size and relative testis size?
• Can you think of an alternative explanation
(besides sperm competition)?
29
Phenotypic plasticity as an adaptive strategy
Phototaxis in Daphnia magna (De Meester, 1996) – 1
• Phenotypic plasticity means that the same genotype may
express different phenotypes
• In general, quantitative phenotypes are plastic
• However, the term phenotypic plasticity is generally
reserved for situations where the plasticity increases fitness
(i.e., is adaptive)
• Daphnia magna is a small fresh-water crustacean
• Reproduces parthenogenetically
• Clones vary in their response to light (phototaxis)
• Phototactic behavior may be plastic within clones
• Daphnia from lakes that have fish (visual predators) tend
to be more negatively phototactic when they sense fish
30
Pictures of Daphnia
31
Variation in phototactic behavior of Daphnia from
three Belgian lakes (Fig. 9.17)
Blankaart has many fish; Driehoekvijver has few fish; Citadelpark has no fish
32
Phenotypic plasticity as an adaptive strategy
Phototaxis in Daphnia magna (De Meester, 1996) – 2
• Most clones from the lake with fish have a plastic
phototactic behavior
• Most clones from lakes with few or no fish are not plastic
• Therefore, phenotypic plasticity (for phototaxis) is itself a
trait that can evolve (there is genetic variation for plasticity
within and among lakes)
• A reasonable hypothesis is that the plasticity of phototaxis
in clones from the lake with fish is an adaptation to reduce
predation by fish
• How do you feel about the level of replication in this
experiment?
• Can you design a stronger test of the hypothesis that plastic
phototaxis is an adaptation to fish predation?
33
Limits to adaptation
• Trade-offs
• Constraints
• Lack of necessary genetic variation
34
Selection on female flower size in Begonia involucrata:
A trade off – 1
• Sexes in separate flowers (monoecious). Male and female
flowers similar in size and appearance
• Bees tend to avoid female flowers (which don’t give a
reward) in favor of male flowers
• But female flowers need bee visits (and pollen) in order to
set seed. Therefore, bees are a strong selective force on
female flower function
• What is the nature of selection imposed by bees on female
flower size?
– Stabilizing selection — best strategy is to look like an average male
flower
– Directional selection — if larger male flowers offer larger rewards then
larger female flowers should be preferred by bees
35
Selection on female flower size in Begonia
involucrata (Fig. 9.19) (Agren and Schemske 1991)
36
Flowers of Begonia involucrata
Fig. 9.18a. Male flower
(left) and female flower An inflorescence
37
Selection on female flower size in Begonia involucrata:
A trade off – 2
• Bees strongly prefer to visit larger flowers
• This is directional selection for larger flowers (male as
well as female)
• What keeps flower size from increasing?
• There is a trade-off between between flower size and
number of flowers per inflorescence
• This suggests that there is net stabilizing selection on
flower size
• What is the optimum balance between flower size and
flower number? How does the plant maximize total seed
set? Indeed, does a plant appear to maximize total seed
set?
38
Flower color change in Fuchsia excorticata: a
constraint – 1
• A bird-pollinated tree
• Flowers are green (5.5 days), then turn red over a period of
1.5 days, then remain on tree as red (5 days) (= 11 days
total)
• Nectar is present during days 1 – 7
• 90% of pollen exported by day 7
• Why do flowers turn red?
– Cue to pollinators, which increases pollination efficiency
• Why bother to turn red, why don’t the flowers just drop off
after 7 days?
39
Flower color change in Fuchsia excorticata (Fig. 9.20)
40
Flower color change in Fuchsia excorticata: a
constraint – 2
• The constraint that appears to prevent dropping of flowers
before they turn red is the time required for pollen tube
growth from the stigma to the ovary.
Table 9.3 Pollen tube growth in Fuchsia excorticata (after
Delph and Lively 1989)
Days since pollination 1 2 3 4
Percentage of 10 flowers with
0 20% 100% 100%
pollen tubes in ovary
41
Host shifts in an herbivorous beetle: constrained by
lack of genetic variation? (Futuyma et al. 1995) – 1
• Herbivorous leaf beetle in genus Ophraella
• Each species feeds on one or several closely related species
of composites (sunflower family, Asteraceae)
• For beetles, the ability to live on a particular host plant
represents a complex adaptation that includes the ability to
recognize a plant as a suitable host and the ability to deal
with plant’s chemical defenses
• What determines the pattern of host plant use by beetles?
Why do closely related beetles not use the same host plant?
42
Phylogeny and host plant relationships of the leaf beetles,
genus Ophraella (Fig. 9.21)
43
Host shifts in an herbivorous beetle: constrained by
lack of genetic variation? (Futuyma et al. 1995) – 2
• Hypothesis 1: all host shift are genetically
possible. The actual distribution of beetles over
host plants is due to ecological or chance factors
• Hypothesis 2: Most host shifts are genetically
impossible. Most beetle species lack sufficient
genetic variation in their feeding preferences and
detoxifying mechanisms to be willing to feed on
and/or to survive on alternate host plants
44
Host shifts in an herbivorous beetle: constrained by
lack of genetic variation? (Futuyma et al. 1995) – 3
• Estimated genetic variation in beetles for feeding on or surviving on
novel hosts using 4 beetle species and 6 host plant species (no genetic
variation = no potential for adaptation to new host)
– no evidence of genetic variation in 18 of 39 tests of whether adults or
larvae would recognize and feed on a potential host
– no evidence of genetic variation for larval survival on potential hosts in 14
of 16 tests
• These results suggest strong genetic constraints on host-plant shifts
• If hypothesis 2 is correct, we can also predict that beetles will be more
likely to show genetic variation for host plant feeding and survival
when potential hosts are closely related to the beetle’s actual host, or
are hosts of closely related beetles (see Table 9.4)
– Genetic variation for feeding in 7/8 tests when novel plant is in same tribe
as beetle’s actual host, but only 14/31 when plant is in different tribe
– Genetic variation for feeding in 12/16 tests when novel plant is host to a
beetle in same major clade as tested beetle, but only 9/23 tests when plant
is host to a beetle in a different major clade 45
Selection operates on different levels
• In our example of HIV, we noted that at the level of virions
within individual hosts, selection favored virions that could
evade the host immune system, as well as develop
resistance to drugs – in other words, virions that end up
killing the host individual
• At the level of virions in the entire host population,
however, the best long-term virion strategy might be to
reduce host mortality – because causing exinction of the
host population would be a bad strategy for HIV
• Would it be possible for natural selection on HIV to favor
reduced host mortality?
46
Selection operates on different levels:
mitochondria in yeast - 1
• Non-respiratory (parasitic) mitochondria replicate faster
than normal mitochondria that carry out aerobic respiration
• Yeast cells with normal mitochondria replicate faster than
yeast with non-respiratory mitochondria
• Selection on mitochondria operates on two levels:
– The level of the population of mitochondria within an individual
yeast cell – where non-respiratory mitochondria ―win‖ because they
replicate faster
– The level of yeast cells within a culture – where normal
mitochondria may ―win‖ because the yeast cells that contain them
replicate faster
• Which level wins depends upon the relative strengths of
selection
47
Selection operates on different levels:
mitochondria in yeast - 2
• Experimental system (Taylor et al. 2002)
• Start populations of yeast that contain a mixture of normal
and parasitic mitochondria in each cell
• Grow yeast at small, medium, and large population size
• Small population size = weak selection at the level of yeast
cells within culture: favors parasitic mitochondria
• Large population size = strong selection at the level of yeast
cells within culture: favors normal mitochondria because
they increase the fitness of the yeast cells that carry them
• Determine the proportion of cells that had only parasitic
mitochondria after 150 generations.
48
Selection on cells in populations versus selection on
mitochondria in cells (Fig. 9.22)
49
Selection operates on different levels:
mitochondria in yeast - 3
• Note: an important difference between these HIV and
mitochondria examples is that mitochondria are transmitted
vertically only. (HIV is most commonly transmitted
horizontally.) This means that the fitness of the host cell or
individual is crucial to the fitness of mitochondria (at least
when selection is acting on the host), but not to the fitness
of HIV
50
Analysis of adaptation: Summary
• Hypotheses about the adaptive value of a trait
should be tested
– Experimental manipulation, observation, comparative
method
• There are limits to adaptation
– Trade-offs (pleiotropy), constraints, lack of genetic
variation
• A complete understanding of the adaptive value of
a trait may require analysis of selection on more
than one level
51