Miniature worlds:

1
Miniature worlds Bromeliad food webs as a model
system for ecology
Diane Srivastava
2
(No Transcript)
3
The idea of the archetype If we have a precise
knowledge of that which constitutes the typical
structure of each of these groups, we shall have,
so far, an exhaustive knowledge of the Animal
Kingdom. - T.H. Huxley (1869)
4

• Easily manipulated
• Many replicates possible
• Quick response time

5
?

• Easily manipulated
• Many replicates possible
• Quick response time

6

• Real ecosystem, co-evolved species!
• Difficult to manipulate
• Low replication
• Slow response time

7
?
General ecological principles

• Real ecosystem, co-evolved species!
• Difficult to manipulate
• Low replication
• Slow response time

8
?
Replication vs. realism -David Schindler
9
Container habitats
10
(No Transcript)
11
c. William H. Bond
12
Bromeliad food web
Intermediate predators
Microbial Food web
c. William H. Bond
Bacteria, fungi
Detritus
13
Rotifers

Ciliates
Flagellates
Bacteria, fungi
14
Why bromeliads are useful systems

• Discrete, simple food webs
• Number of trophic levels (with M. Melnychuk,
  J. Ware)

15
Why bromeliads are useful systems

• Discrete, simple food webs
• Stable manipulations of community structure
• Effect of habitat complexity on trophic
  cascades

16
Why bromeliads are useful systems

• Discrete, simple food webs
• Stable manipulations of community structure
• Scale of population dynamics differs with taxa
• Extinction cascades (with T. Bell)

17
Why bromeliads are useful systems

• Discrete, simple food webs
• Stable manipulations of community structure
• Scale of population dynamics differs with taxa
• Similar habitat occurs over broad geographic
  range
• Biogeographical comparisons (B. Richardson)
• Contained ecosystem for nutrient budgets
• Insects and bromeliad growth? (A. Reich, J.
  Ngai)

18
Does energy limit the number of trophic levels?
19
Theory
Energy is lost in the transfer between trophic
levels (about 10 transfer efficiency). If
energy is limiting, trophic diversity will be a
logarithmic function of basal energy (every 10x
increase in energy can support one more trophic
level).
20
Theory
Energy is lost in the transfer between trophic
levels (about 10 transfer efficiency). If
energy is limiting, trophic diversity will be a
logarithmic function of basal energy (every 10x
increase in energy can support one more trophic
level).
Difficult to quantify basal energy, number of
trophic levels
21

• 10x increase in energy correlated with lt 1
  trophic level
• Intraguild predation will decrease trophic
  levels
• Covariates?

22
(No Transcript)
23
No damselfly larvae below 100 ml capacity
24
1997
2000
Bromeliad capacity (ml)
25
1997
Prey available (g per M. modesta larva)
2000
Bromeliad capacity (ml)
26
Bromeliad capacity (ml)
27
Bromeliad capacity (ml)
28
(No Transcript)
29
Effect of habitat complexity on trophic cascades
30
Theory
Trophic cascades occur when there are strong
linear trophic links. Habitat complexity may
increase predator search time, or increase prey
survival (refuges)
31
Theory
Trophic cascades occur when there are strong
linear trophic links. Habitat complexity may
increase predator search time, or increase prey
survival (refuges)
Manipulating habitat complexity, isolating
effects on predators and prey
32
Experimental design
Top trophic level
33
Experimental design
Top trophic level
X
Bromeliad complexity
34
Experimental design
Top trophic level
X
Bromeliad complexity
X
Bromeliad size (Expt. 2 only)
35
Expt. 1


Predation x complexity or complexity2 Plt0.05
36
Expt. 1


Predation x complexity Plt0.0001
37
Insects grow more slowly in complex bromeliads
38
Effect of predator diminishes with complexityand
size
Predator x Complexity Plt0.05
Detrital processing No predator - Predator
1
3
6
Complexity
Larger bromeliads also have reduced foraging
efficiency
Detrital processing Small - Large
1
3
6
Complexity
39
Effect of predator diminishes with complexityand
size
Detrital processing No predator - Predator
1
3
6
Complexity
Larger bromeliads also have reduced foraging
efficiency
Detrital processing Small - Large
1
3
6
Complexity
40
Effect of predator diminishes with complexityand
size
Predator x Size Plt0.05
Detrital processing No predator - Predator
1
3
6
Complexity
Larger bromeliads also have reduced foraging
efficiency
Detrital processing Small - Large
1
3
6
Complexity
41
Effect of predator diminishes with complexityand
size
Predator x Size Plt0.05
Detrital processing No predator - Predator
1
3
6
Complexity
Larger bromeliads also have reduced foraging
efficiency
No predator Size effect P0.01 Predator NO
Size effect P0.88
Detrital processing Small - Large
1
3
6
Complexity
42
Increased bromeliad complexity
- Decreased detritivore efficiency
Direct effect
Effect on detrital processing
43
(No Transcript)
44
(No Transcript)
45
Alex Reich
Bromeliad growth experiment
46
What happens to food webs and ecosystems when
species go extinct?
47
Theory

• Declining species diversity will cause
• Loss of species at lower trophic levels
  (extinction cascades)
• Reduction in ecosystem functions

48
Theory

• Declining species diversity will cause
• Loss of species at lower trophic levels
  (extinction cascades)
• Reduction in ecosystem functions
• Manipulating animal diversity!

Problem
49
Experimental design
Top predator
extinction
response (cascade)
Detritivores
Ciliates
extinction
Detritus
response (function)
50
Detritivore communities
Red chironomid (R)
Yellow chironomid (Y)
Tipulid (T)
Helodid (H)

• 1 species (4 community types) T H R
  Y
• 2 species (6 community types) TH TR TY HR
  HY RY
• 4 species (1 community type) THRY

51
All communities are designed to have,
theoretically, the same metabolic
capacity Therefore, differences amongst
communities are true effects of composition
T H TH THRY
52
Are there extinction cascades?
53
Ciliate species richness
TR
No damselfly
TY
TH
R
HY
Y
THRY
RY
HR
No insects
T
H
H, HY, TH, TR, TY plt0.05
54
Ciliate species richness
TR
TY
TH
R
HY
Y
THRY
No damselfly
RY
HR
T
H
H, HY, TH, TR, TY plt0.05
55
Ciliate species richness
Full model (F tests, plt0.05) Species identity
Species interactions (TH, TR plt0.05) Trophic
diversity Trophic diversity x species interactions
TR
TY
TH
R
HY
Y
THRY
No damselfly
RY
HR
T
H
H, HY, TH, TR, TY plt0.05
TR
With damselfly
TH
Y
R
HR
H
THRY
RY
HY
TY
T
56
Does insect diversity affect decomposition?
Decomposition
Diversity of insects
57
Decomposition
T
TY
No damselfly
TH
TR
HR
H
THRY
HY
R
RY
Y
No insects
H,R,T,Y,TY plt0.05
58
Decomposition
Full model Species identity (T,H,R,Y) Species
interactions Trophic diversity (plt0.05) Trophic x
species interactions
No damselfly
H,R,T,Y,TY plt0.05
59
Decomposition
Full model Species identity (T,H,R,Y) Species
interactions Trophic diversity (plt0.05) Trophic x
species interactions
No damselfly
H,R,T,Y,TY plt0.05
With damselfly
60

Rotifers
Ciliates
Flagellates
Bacteria, fungi
Detrital processing chain
Fine particles
Coarse particles
61
ALL SIGNIFICANT SPECIES INTERACTIONS (No
damselfly bromeliads)
Detrital loss Ciliate richness Ciliate density Flagellate density Rotifer density
HR
HY
RY
TH
TR
TY
THRY
62

• Summary
• Evidence of indirect extinction cascades between
  insects and ciliates, possibly due to processing
  chains
• Decomposition is more strongly affected by
  vertical (trophic levels) than horizontal
  (species) extinctions

63
Could container habitats be model systems for
ecology?
Simple quantifications of habitat size,
complexity and basal energy Easy manipulations
of diversity and trophic structure Quantifiable
food webs and ecosystem functions Real
co-evolved communities!
64
Flagellate density
Rotifer density
Ciliate density
H
No damselfly
H
H
H, TH, TR plt0.05
H, HR plt0.05
H plt0.05
H
H
With damselfly
H
Full model Species identity Species
interactions Trophic diversity Trophic x species
interactions
Full model Species identity (H p0.003) Species
interactions Trophic diversity Trophic x species
interactions
Full model Species identity (H plt0.001) Species
interactions Trophic diversity Trophic x species
interactions
65
Interaction regressions 1. Account (test) for
species identity effects. tip hel red yel
T 1 0 0 0 H 0 1 0 0 TH 0.5 0.5 0 0 THRY 0.
25 0.25 0.25 0.25
Function ß0ß1tip ß2hel ß3red ß4yel
Regression term
Treatment
66
Interaction regressions 1. Account (test) for
species identity effects. 2. Test if
multi-species communities have functions
different than expected from the sum of their
parts (species interactions) tip hel red yel
tiphel all T 1 0 0 0 0 0 H 0 1 0 0 0 0 TH 0.5
0.5 0 0 1 0 THRY 0.25 0.25 0.25 0.25 0 1
Regression term
Treatment
Function ß0ß1tip ß2hel ß3red ß4yel
ß5tiphel
67
Interaction regressions 1. Account (test) for
species identity effects. 2. Test if
multi-species communities have functions
different than expected from the sum of their
parts (species interactions) tip hel red yel
tiphel all T 1 0 0 0 0 0 H 0 1 0 0 0 0 TH 0.5
0.5 0 0 1 0 THRY 0.25 0.25 0.25 0.25 0 1 3.
Cross everything with trophic diversity
Regression term
Treatment