These are the slides from a talk I gave about how pollinator foraging behavior can affect flowering plant coexistence. This talk was in the Species Interactions III group at the Ecological Society of America Conference in Portland on August 10, 2017. Included: bonus slides!

1. The effects of pollinator sharing on
flowering plant coexistence
Aubrie James and Monica Geber
Cornell University

3. What does pollinator sharing mean for
flowering plant coexistence?

4. Case Study:
Clarkia in the
Kern River
Canyon
C. speciosa C. xantiana
C. unguiculataC. cylindrica

5. Case Study:
Clarkia in the
Kern River
Canyon
C. speciosa C. xantiana
C. unguiculataC. cylindrica

6. Case Study:
Clarkia in the
Kern River
Canyon
C. speciosa C. xantiana
C. unguiculataC. cylindrica
Moeller 2004, Ecology

7. ↵How should we think about pollinators
affecting Clarkia coexistence?
Quantify pollinator preference
Quantify pollinator constancy
1. Generate predictions based on pollinator foraging behavior
2. Ask if bee foraging behavior explains C. xantiana facilitation

8. Modern Coexistence Theory

9. Modern Coexistence Theory
Negative frequency
dependence
“effect on own species”
Frequency in the community
Percapitagrowthrate
Figure after Adler et al 2007

10. Modern Coexistence Theory
Negative frequency
dependence
“effect on own species”
Frequency in the community
Percapitagrowthrate
Figure after Adler et al 2007
Frequency in the community
Percapitagrowthrate
“effect on others”
Fitness inequality

11. Do shared pollinators change
fitness inequality or negative
frequency dependence?
Pollinator
Constancy
Pollinator
Preference
1. Generate predictions based on pollinator foraging behavior

12. Pollinator
preference =
fitness
inequality
PreferencePreference
Pollinator preferences introduce fitness inequality

13. Foraging constancy can change
negative frequency dependence
Inconstant
foraging =
destablizing,
punishes rare
species for
being rare
constant bee inconstant bee
frequency in communityfrequency in community
percapitagrowthrate

14. Have to determine how pollinators forage in
different patch types
vary floral abundance in an array
vary floral diversity in an array

15. Experimental Design
2015 2016
C. xantiana
• Four diversity treatments
• Four density treatments
• Four diversity treatments
• Four density treatments

16. Pollinator preferences can introduce fitness inequality
Frequency in the community
Percapitagrowthrate
Fitness inequality
Preference

17. Pollinator preferences introduce fitness inequality
Frequency in the community
Percapitagrowthrate
Fitness inequality
Preference
Frequency in the community
Percapitagrowthrate
Fitness inequality
Preference

18. -0.50
-0.25
0.00
0.25
0.50
C S U X
Species
PollinatorPreference
In both years:
Pollinators
introduce
fitness
inequality
between C.
xantiana and
the other
Clarkia species
Pollinators prefer Clarkia xantiana
XUSC
Preference for i = (Proportion of visits to i Proportion of i in array)

19. Lesson 1: Sharing pollinators
introduces a fitness
inequality because of
pollinators’ preference for
Clarkia xantiana
(or, don’t hang out with your super hot
friend if you’re trying to get a date)

20. Foraging constancy can change
negative frequency dependence
frequency in community
percapitagrowthrate
Constancy

21. Foraging constancy can change
negative frequency dependence
frequency in community
percapitagrowthrate
Constancy
frequency in community
percapitagrowthrate
Constancy

22. 0.00
0.25
0.50
0.75
1.00
0.4 0.6 0.8 1.0 1.2 1.4
Array Diversity
PollinatorConstancy
Pollinator constancy declines with array diversity
In both years:
constancy
declines with
diversity
Interspecific
pollen transfer
more likely in
diverse arrays
Shannon’s Diversity in Array
AverageConstancy

23. Lesson 2: Sharing
pollinators is
destabilizing in diverse
Clarkia patches because
pollinators are inconstant
(or, there’s safety in numbers)

24. ↵How should we think about pollinators
affecting Clarkia coexistence?
Quantify pollinator preference
Quantify pollinator constancy
1. Generate predictions based on pollinator foraging behavior
2. Ask if bee foraging behavior explains C. xantiana facilitation

25. ↵How should we think about pollinators
affecting Clarkia coexistence?
2. Ask if bee foraging behavior explains C. xantiana facilitation
Quantify pollinator preference
Quantify pollinator constancy
1. Generate predictions based on pollinator foraging behavior
Pollinator preference for C. xantiana =
fitness inequality
Could promote or hinder coexistence

26. ↵How should we think about pollinators
affecting Clarkia coexistence?
2. Ask if bee foraging behavior explains C. xantiana facilitation
Quantify pollinator preference
Quantify pollinator constancy
1. Generate predictions based on pollinator foraging behavior
Pollinator preference for C. xantiana =
fitness inequality
Could promote or hinder coexistence
Constancy decreases with diversity,
destabilizes diverse plant assemblages
Makes coexistence harder

27. Bees are highly mobile, choosy, and select
abundant, diverse patches to forage in
2. Ask if bee foraging behavior explains Clarkia facilitation
Facilitation via joint attraction

28. Facilitation via joint attraction
Holling’s Functional Responses
Density of Resource
ConsumptionofResource
Holling 1959 Can. Entomol.
Hilborn 2015 Advances in
Ecological Research

29. Facilitation via joint attraction
N =
Nmax
✓
P
i
Fi
◆
+
✓
P
i
Fi
◆
Parameter Definition
maximum visitation
rate (visits/bee)
visitation rate (visits/
bee)
sum of all species
(array density)
determines if
function
decelerates or
accelerates
half saturation
constant
N
Nmax
X
i
Fi
Holling 1959 Can. Entomol.
Feldman et al 2004 Oikos
Holling’s Functional Responses

30. Array Density
Visits/Bee
0 1 2
012345
NmaxDensityσ
β + Densityσ
σ = 0.75
σ = 2
Facilitation via joint attraction
0 <  1
1 <
Type II, Saturating, No Facilitation
Type III, Accelerating, Facilitation
N =
Nmax
✓
P
i
Fi
◆
+
✓
P
i
Fi
◆

31. Array Density
Visits/Bee
100 200 300 400 500
12345
If co-occurring with other flowers is facilitative via the joint
attraction of pollinators, visitation rate will accelerate with array
density
Parameter Definition
ymax
maximum
visitation rate
(visits/bee)
y
visitation rate
(visits/bee)
x array density
determines if
function
decelerates or
accelerates or
decelerates
half saturation
constant
nls
✓
y ⇠
ymaxx
+ x
◆

32. Parameter Estimate
0.19 ± 0.05
5.5e6 ± 1.5e12
Visitation rate decelerates with array density; facilitation not a
result of bee foraging behavior
nls
✓
y ⇠
ymaxx
+ x
◆
Array Density
Visits/Bee
0 100 200 300 400 500
12345
NmaxDensityσ
β + Densityσ

33. Lesson 3: Shared pollinator
foraging behavior does not
have a facilitative effect on
co-blooming plants
(or, everyone slows down at a buffet)

34. ↵How should we think about pollinators
affecting Clarkia coexistence?
2. Ask if bee foraging behavior explains C. xantiana facilitation
Quantify pollinator preference
Quantify pollinator constancy
1. Generate predictions based on pollinator foraging behavior
Pollinator preference for C. xantiana =
fitness inequality
Could promote or hinder coexistence
Constancy decreases with diversity,
destabilizes diverse plant assemblages
Makes coexistence harder
No, because visitation rate slows down!
Something else is going on here

35. Thank you!
Talk to me!
email: aj465@cornell.edu
twitter: @slaubrie
website: www.aubriejames.com
Dr.Monica Geber
Dr. Anurag Agrawal
Dr. Steve Ellner

40. Facilitation
Two ways to get there:
1. More and more reliable
pollinators in diverse plant
communities (Numerical
response)
2. Bees are highly mobile,
choosy, and select abundant,
diverse patches to forage in
(Behavioral response)

41. Array Density
Visits/Bee
0 100 200 300 400 500
12345
NmaxDensityσ
β + Densityσ
Array Density
Visits/Bee
0 1 2
012345
NmaxDensityσ
β + Densityσ
σ = 0.75
σ = 2
σ = 0.19 ± 0.05
Visitation rate decelerates with array density; facilitation not a
result of bee foraging behavior

42. Constancy changes according to array type
0.6
0.7
0.8
0.9
1.0
XC XCSU XS
ArrayType
Constancy0.5
0.6
0.7
0.8
0.9
XC XCSU XS
ArrayType
Constancy
2015 2016

43. 0.0
0.5
1.0
1.5
100 200 300 400 500
Number of Flowers
Beesenteringarray/minute
Species in Array
X
XC
XCSU
XS

44. Bateman’s index
stancy =
(AD)1/2
(BC)1/2
(AD)1/2 (BC)1/2
A B
C D
Move1
Move 2
Constancy =
(Diagonal)1/4
(OffDiagonal)1/12
(Diagonal)1/4 (OffDiagonal)1/12
Move1
Move 2

45. 1. Behavioral Constancy =/=
foraging constancy! Monitoring
behavior doesn’t necessarily give us estimates
of pollen constancy and pollen preference

46. 1. Behavioral Constancy =/=
foraging constancy! Monitoring
behavior doesn’t necessarily give us estimates
of pollen constancy and pollen preference
-0.4 -0.2 0.0 0.2 0.4
-0.40.00.2
SNP PC1 (10.5%)
SNP PC2
(8.4%)
SNP PC
(6.9%)
S
C
XU

47. Are bees
constant? yes
no
Do bees
prefer the
competitive
dominant?
no
Prediction: pollinators
destabilize
plant communities
Prediction:
pollinators
decrease plant
fitness differences
Prediction:
pollinators increase
plant fitness
differences
yes

48. y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413y = 1 − 0.31 ⋅ x, r2
= 0.413
0.00
0.25
0.50
0.75
1.00
0.0 0.5 1.0
Diversity
Constancy
Number of Species
1
2
4
2015 Arrays
1. Generate predictions based on pollinator foraging behavior
Are bees
constant?
no
Prediction: pollinators
destabilize
plant communities
Test: are they
inconstant enough
to matter?
• Measure interspecific
pollen transfer, pollen
limitation for rare plants
• Measure reproductive
performance when rare

49. 1. Generate predictions based on pollinator foraging behavior
Are bees
constant?approximately
Do bees
prefer the
competitive
dominant?
Prediction:
pollinators increase
plant fitness
differences
yes…
…(if) the
competitive
dominant is
Clarkia xantiana
Test: Do bee preferences change
fitness differences in Clarkia?
• Measure bee visitation, reproductive
performance when not rare and
compare across species