Title: Foraging Causes and Consequences of Consumer Behavior
1ForagingCauses and Consequences of Consumer
Behavior
- Peter B. McEvoy
- Insect Ecology
- Ent 420/520
2Consumer Behavior
- Physiological ecology how organisms respond to
food (e.g. visual, chemical cues) - Behavioral ecology how natural selection has
favored particular patterns of consumer behavior
in particular environments (e.g. optimal
foraging) - Populations dynamics how behavior of predator
and prey influence their population dynamics
3Consumer Searching Behavior Contrasting Views
and Approaches
- Contrast nonrandom laboratory view and random
field view (Morris and Kareiva 1991) - Contrast phenomenological and mechanistic
approaches (Turchin 1998)
4Physiological Ecology of Host-plant Selection A
Catenary Process
- Behavioral elements follow each other in fixed
order or reaction chain - random movement
- searching guided by olfactory and/or optical cues
- contact/evaluation using visual, olfactory,
mechanosensory, gustatory information - acceptance/rejection
- What cues give reliable information at each step?
5Generalized Sequence of Events in Host-plant
Selection
Schoonhoven et al 1998
6Behavioral Ecology of Consumer-Resource Relations
- Consumer Diets (specialization)
- Individual Behavior (functional, numerical,
aggregative responses) - From Individual Behavior to Population Dynamics
(aggregation of risk in relation to resource
density) - Optimal Foraging Approach (critique of the
adaptationist program)
7Responses of consumer to changes in resource
density
- Behavioral response within generation
- Functional response is a change in per capita
consumption rate of consumer brought about by
change in resource density - Congregative response is movement of consumer
leading to concentration of consumers in areas of
high resource density - Reproductive response between generations
- Numerical response is a change in per capita
reproductive rate of consumer in response to
change in abundance of resource
8Representing consumer responses in models
9Food Preferences of Consumers
- Food preferences. Necessary to examine both food
in diet and availability of different food
types in the environment. Availability must be
seen not through the eyes of the observer but
through the eyes of the insect itself. - Ranked and balanced preferences. Ranked
preferences predominate when food items can be
classified on a single scale (many carnivores). - Mixed diets. Many consumers exhibit a
combination of ranked and balanced preferences
(many herbivores). Performance is far better on a
mixed diet than on a pure diet of even the best
food. - Switching. In contrast to a fixed preference,
switching involves a preference for the more
frequent food types.
10Concept Alert!
- Consumer preference vs performance. Resource use
reflects both consumer choice (preference) and
consumer ability to convert resource into useful
outputs like growth and reproduction (performance)
11Host Specializationin Swallowtail butterfly
Papilio machaon
12Host SpecializationPreference and Performance in
Papilio machaon
Larval performance Most plants are suitable food
plants, but females lay eggs on only a fraction
Hierarchy of oviposition preferences Females lay
eggs on plants that do not support growth
13Oviposition by Papilio machaon occurs on plants
unsuitable for larvae
- Two invasive plant species, Bifora radians and
Levisticum officinale, both Apiaceae
(Umbelliferae), were introduced into Scandinavia
and are rare oviposition behavior has not yet
evolved to avoid these species
Bifora radians
Levisticum officinale
14Host range in taxonomically conservative
- Plants in subfamily Apiodeae suitable for larvae,
those in Hydrocotyloideae and Saniculoideae
resulted in 100 mortality
Saniculoideae (Sanicula, Eryngium),
Hydrocotyloideae (Hydrocotyle).
http//biodiversity.uno.edu/delta/angio/www/umbell
if.htm
15Host range in Lab gtgtHost Range in Field
Papilio machaon
Angelica archangelica ssp. littoralis
- Total of 33 spp. Apiaceae used in lab, only 1 sp
(Angelica archangelica ssp. littoralis) used in
field
16Summary Papilio machaon study
- Invasive plant species. Oviposition on hosts
unsuitable for larvae. Bifora radians and
Levisticum officinale, both Apiaceae
(Umbelliferae), were introduced into Scandinavia
and are rare oviposition behavior not yet
evolved to avoid these species - Host Range Taxonomically Conservative Plants in
Apiodeae suitable for larvae, those in
Hydrocotyloideae and Saniculoideae resulted in
100 mortality - Host range in Lab gtgtHost Range in Field. Total
of 33 spp. Apiaceae used in lab, only 1 sp
(Angelica archangelica ssp. littoralis) used in
field - Unsettled points. Preference and performance in
larvae, abiotic and biotic factors in wild may be
absent from the lab, host a habitat as well as
food
17Pattern of Movement Behavior
- Key behavioral characteristics
- Frequency, the proportion of time spent moving
- Rate, the rate of movement when actually in
motion - Orientation, the direction of movement
- Three forms of movement
- Nonrandom settlement, to move or not to move,
that is the question rate and orientation while
moving are random - Area restricted search refers to change in
consumer searching behavior (increase in turning
frequency and/or decrease in velocity) following
feeding - Orientation, consumer uses cues regarding
neighboring plants (odors or visual cues) to
alter direction of movement
18Insect Movement in Heterogeneous Space (Turchin
and Omland 1999)
- Distribution of food (directed movement or taxis,
undirected movement or kinesis) - Tracking abiotic factors
- Distribution of conspecifics (DD congregation,
dispersal, mate finding) - Avoiding predation
- Dealing with multiple factors
19How Herbivores Use Odors to Orient Toward Host
Plants Experimental Setup
20Orientation to Upwind Odor (External Stimuli)
Mediated by Satiation (Internal State)
Male
Female
Orientation when Starved
Random when Satiated
Ragwort flea beetle Longitarsus jacobaeae
21Taxis Directed Movement in Response to
OdorColorado Potato Beetle
(B) Pure tomato
(A) Clean air
(D) Mix tomato potato Associational resistance
(C) Pure potato
Note wind direction, mean vector length and
upwind component
22Analysis of Movement Pathsstep length/time
(velocity), turning rate, turning angle
23House Flies Tend to Remain Where Resources Are
Concentrated
Large turning angle, small step length
immediately after feeding, then return to ground
state
24Bumblebees Use Area Restricted Search
Low nectar, longer moves, lower turning angle
Bumblebees use Area Restricted Search
High nectar, shorter moves, higher turning angle
25Patch Residence Time in Colias Regulated by
Changing Step Length
Median percent cover
Median step length
Median step length
26Aphis varians Congregate by Increasing
Probability of Settling With Increasing Aphid
Density
27Functional Responses Consumption Rate X Food
Density
- Type 2 functional responses and handling time
- Parasitoid Pleolophus basizonus attacking cocoons
of the European pine sawfly Neodiprion sertifer - Type 2 response can be defined by handling time
and searching efficiency. Hollings disc
equation - Alternative reasons for a type 2 functional
response - Type 1 functional response
- Type 3 functional response will arise whenever an
increase in food density leads to an increase in
the consumers searching efficiency or decrease
its handling time
28Consequences of Functional Responses for the
Dynamics of Populations
Does functional response translate into direct
density dependence in mortality rate?
Prey killed per predator
Mortality Rate
Prey density
Prey density
29Effects of Consumer Density Mutual Interference
Leads to Reduction in Consumption Rate
- Coefficient of interference plotting log10
searching efficiency as a function of consumer
density. Slope takes the value of m and m is
known as the coefficient of interference - The reverse of interference is facilitation
- Mutual interference tends to stabilize
predator-prey dynamics
30Consumers, Food Patches, and Aggregative Responses
- Eye of the beholder. Patch must be defined with a
particular consumer in mind - Definition. Patch is area within which
homogeneity can be assumed - Food value. Patches can vary in the density of
food or prey they contain - Predation risk. Risk of predation often varies
among patches differing in density
31Aggregation of Risk Traditional View
- Some simple expectations
- Allocation of time. Consumers generally spend
more time in patches containing high densities
(because these are the most profitable patches) - Distribution of consumers. Most consumers are
therefore found in such patches - Risk of being consumed. Prey in those patches
are therefore more vulnerable to predation
whereas those in low density-patches are
relatively protected and more likely to survive - Often contradicted by observations responses may
be directly or inversely density dependent,
domed, or density independent - For herbivores (Cromartie 1975 Kareiva 1983)
- For parasitoids (Lessells 1985, Stiling 1987,
Walde Murdoch 1988, Pacala Hassell 1991)
32Aggregation of Risk Modern View
- Causes of clumping. Herbivores often aggregate
without this being an aggregative response - Aggregation of risk. Heterogeneity in the risk of
being consumed may reflect individual variation
in resistance, timing, or spatial location
33How Flea Beetles Forage Among Patches Differing
in Plant Density
- Organisms. Flea beetle Longitarsus jacobaeae
foraging among patches of its host plant ragwort
Senecio jacobaea - Methods. Employed mass mark-recapture methods
because it is difficult to follow individuals - Processes. Study includes colonization, birth,
death, and interaction components - Time scale assumptions. Assume colonization and
redistribution process occurs on faster time
scale than birth, death, and population
interactions. Organisms move, then interact. - Aggregation mechanisms. Diffusive, kinetic, and
taxis mechanisms allow individual foragers to
regulate their residence time in resource patches
according to the quality of the patch
34Ragwort Biocontrol
35Life Cycle of Ragwort Flea Beetle
36Driving Forces
- Disturbance removes organisms, recycles
limiting resources, sets the stage for
colonization and occupancy - Colonization movement to the resource mediated
by random and nonrandom elements - Local Interactions plant (insect) competition
and insect herbivory
37Equilibrium Density of Foragers
Evaluated at equilibrium, proportional increase
in hosts causes lt proportional increase in
beetles (slope lt 1)
Beetle population reached equilibrium day 7
38Colonizing Abilities of 3 Insects Introduced for
Ragwort ControlProportion of patches occupied x
distance from source
Flea Beetle lt
Cinnabar moth lt
Seed-head fly
Proportion
Distance from Source
D50 Distance with probability of 50
infestation rate at origin
39Year to Year Variation in ColonizationProportion
of Plants Occupied X Distance From Source
D50 148 m
D50 119 m
Proportion declines with distance in one year
Proportion independent of distance the next year
D50 Distance with probability of 50
infestation rate at origin
40Factors Influencing ColonizationRagwort Flea
Beetle
- Temperature Cold snap (lt 5oC)
- Season Rise and fall of beetle numbers during
colonizing phase - Hosts per patch Amplitude of colonization curve
increases with the number of hosts per patch
(host area) - Distance from source Amplitude of colonization
curve grows weaker with distance from source
41Flea Beetles Emerging the Following Year
- Amplitude of beetle emergence curve increases
with number of hosts per patch - Cumulative number of beetles emerging increases
with number of hosts per patch
42Growth Rate of Beetle Population
Allee Effect Decrease in growth rate at low
density
N(t1) / N(t)
Male-bias sex ratio in 1-plant patch, female
bias in larger patches
Males/Females
43Growth Rate of Plant PopulationHerbivore Effect
Independent of Plant Density
Beetles excluded
Beetles present
44Summary of Flea Beetle Study
- Evidence of consumer/resource equilibrium
encourages focus on equilibrium rather than
transient states - Causes of aggregation highlight mix of random and
nonrandom behavior (diffusion, kinesis, taxis) - Consequences of aggregation for growth rates of
consumer and plant populations, dynamics of
interaction. No refuge from parasitism due to
heterogeneity of risk.
45Quantitative Analysis of Foraging Movement
(Morris and Kareiva 1991, Turchin 1998, Turchin
and Omland 1999)
- Represent random and nonrandom components of
searching behavior in a mathematical model
(diffusion, kinesis, taxis)
46They Find Empirical Studies Are Not Measuring
the Right Things
- Current state of our knowledge. Empirical studies
provide quantitative measurements of all three
components of movement (diffusion, kinesis,
taxis) in only 2 of 45 species (i.e. mite
Tetranychus urticae and butterfly Pieris rapae) - Consequences of our ignorance Failure to
understand how insects locate superior food
plants hampers investigation of - Host race formation and evolution of diets
- Plant-herbivore coevolution
- Implications of different cropping patterns
- Safety and effectiveness of biocontrol organisms
- Consequences of changing food quality of
herbivore populations for population dynamics
47Key Questions for Empirical StudiesMorris and
Kareiva
- Should I distinguish mobile and sedentary
subpopulations? Should the foraging population be
divided into mobile and sedentary subpopulations,
and can it be treated as one pool of mobile
individuals? - How do I assess sensitivity to variation in host
quality? Does the herbivore alter its rate of
movement in accord with spatial variation in the
quality of its potential host plants? - How do I screen for orientation? Do individuals
orient and move in a directed manner toward
plants whose quality stands above the quality of
surrounding vegetation?
48Discussion
- What processes besides movement can lead to
aggregation of consumers? - How might proponents of foraging theory respond
to criticisms leveled by Lewontin and Gould? - Why are phenomenological models so difficult to
relate to actual movement behavior? - Now that we have have mechanistically-grounded
models, is there any need to employ
phenomenological models lacking such a basis?
49Refuge Theory Now the Outcome of Biocontrol
(Y) Can Be Predicted From a Single, Easily
Measured Parameter (X) (Hold the hype)
Hawkins et al 1994
50Study System
Tachinid parasitoid, Tachinomyia similis of
western tussock moth (Orgyia vetusta Bdv.
Lymantriidae)
Bush lupine (Lupinus arboreus)
51Mechanisms leading to inverse-density dependence
in parasitism
- Leaving patches at a constant rate to avoid
self-superparasitism - Decelerating functional responses caused by
behaviors such as handling time or group defenses - Interference among parasitoids
52Alternative Models of Parasitism
- No effect of host density on parasitism.
- Weekly variability in parasitoid aggregation, no
decelerating functional response. - Decelerating (type 2) functional response
- Decelerating functional response and weekly
variability in parasitoid aggregation
53Maximum Likelihood Method
- Used to fit a regression line to bivariate or
multivariate data rather than the least-squares
method - Will result in values for the parameters that
make the observed values in our data set seem
most probable - Solution is carried out iteratively on computer
54Akaike information criterion
- Criterion for selecting among models
- Best model is one with lowest AIC
- Compromise between minimizing the number of
parameters and minimizing the residual variance - The AIC is a number associated with each model
- AICln (sm2) 2m/T
-
- where m is the number of parameters in the model,
and sm2 is (in an AR(m) example) the estimated
residual variance sm2 (sum of squared
residuals for model m)/T, and T the number of
observations.
http//economics.about.com/cs/economicsglossary/g/
akaikes.htm
55Fig. 1. Foraging Tachinomyia similis (fly)
population response to variation in Orgyia
vetusta (host) density.
56Fig. 2. Attack rates by T. similis in relation
to experimentally manipulated density of O.
vetusta and T. similis.
- Host treatment effect (plt0.001)
- No Fly treatment effect (P0.73)
57Fig. 3. Observed parasitism rates by T. similis
(fly) throughout the season in relation to
densities of larval O. vetusta (host).
- Model 2 (dashed) includes an effect of parasitoid
aggregation - Model 4 (solid) includes effect of both
parasitoid aggregation and saturating functional
response - ß the increase in fly density expected with
increases in host density - s the strength of saturation in the functional
response
58General Application
- maximizing Eq. 5 where the expected percent
parasitism, pj, is described by the general form
1expF(H,P)A(H,P)t where F(H,P) is the
functional response of the predator in units of
(parasitoid density time)-1 and A(H,P) is the
aggregative response of the predator in units of
parasitoid density and t is the length of the
experiment.