Null models and observed patterns of native and exotic diversity: Does native richness repel invasio

1
Null models and observed patterns of native and
exotic diversity Does native richness repel
invasion?

• Rebecca L. Brown,1,2 Jason D. Fridley,1
• and John F. Bruno1
• 1University of North Carolina-Chapel Hill
• 2Patrick Center for Environmental Research

2
Does diversity control invasion?

-
Confounded or neutral relationship

• Prieur-Richard et al. 2000 Stachowitz et al. 2002
  Dukes 2002
• Tilman 1997
• Hector et al. 2001
• Knops et al. 1999
• (mostly experimental)

Stohlgren and Chong 2002 Wiser et al. 1998 Bruno
et al. 2002 Burger et al. 2001 Sax
2002 Lonsdale 1999 (mostly observational)
Levine 2001 Lavorel et al. 1999. Brown and
Fridley 2003 Duncan 1996 Stohlgren et al.
1998 Brown and Peet 2003
3
Spatial scale effects

• Small scales
• Saturation, plant to plant competition
• negative relationship

Larger scales Variation in other factors
(disturbance, propagules, fertility) positive
or no relationship
4
2
2
2
2
100 m
1 m
0.1 m
0.01 m
3
3
4
40
3
30
2
2
2
Exotic species richness
20
1
1
1
10
0
0
0
0
0
2
4
6
8
10
0
1
2
3
4
5
0
5
10
15
20
0
20
40
60
80
100
Native species richness
5
BUT do these relationships imply biological
mechanisms or could they be observed in randomly
assembled communities?

• What is the scale-dependence of the native-exotic
  richness relationship in a randomly assembled
  community?

6
Randomly assembled communities the null model

• Create simulated communities with native and
  exotic species sampled at multiple scales
• Randomize native and exotic species codes in real
  communities

7
(No Transcript)
8
Simulation method
Randomly pick of sp in pool, assign abund. to
each sp
20 100 spp 1-10,000 individuals
Pool 75 native 15 exotic
10 blank
When full, repeat process 100 times for each
quadrat scale
Randomly pick 1 individual from pool
Repeat until quadrat is filled (5 to 800 spaces)
Add one individual of that sp to quadrat
9
Simulation of random quadrats
N 800
Exotic Richness
Native Richness
10
Simulation of random quadrats
N 800
N 100
Exotic Richness
Native Richness
Native Richness
11
Simulation of random quadrats
N 800
N 100
N 50
Exotic Richness
Native Richness
Native Richness
Native Richness
12
Simulation of random quadrats
N 800
N 100
N 50
Exotic Richness
Native Richness
Native Richness
N 20
Exotic Richness
Native Richness
13
Simulation of random quadrats
N 800
N 100
N 50
N 50
Exotic Richness
Native Richness
N 20
N 10
Exotic Richness
Native Richness
Native Richness
14
Simulation of random quadrats
N 800
N 100
N 50
Exotic Richness
N 20
N 5
N 10
Exotic Richness
Native Richness
Native Richness
Native Richness
15
Constraints on high native-exotic richness at
smallest scales
100
Species pool 75 Natives, 15 Exotics
100 individuals
Exotic species richness
50
Null relationship
0
0
50
100
Native species richness
16
Constraints on high native-exotic richness at
smallest scales
100
Species pool 75 Natives, 15 Exotics
100 individuals
Exotic species richness
50
50 individuals
Null relationship
0
0
50
100
Native species richness
17
Constraints on high native-exotic richness at
smallest scales
100
Species pool 75 Natives, 15 Exotics
100 individuals
Exotic species richness
50
50 individuals
Null relationship
10 individuals
0
0
50
100
Native species richness
18
100
Species pool 75 Natives, 15 Exotics
100 individuals
Exotic species richness
50
50 individuals
Null relationship
10 individuals
0
0
50
100
Native species richness
19
Summary Simulated Data

• In simulated randomly assembled communities, the
  relationship between native and exotic richness
  is positive at large scales, and negative at
  small scales
• Positive because plots differ in total richness
  slope is simply ratio of natives to exotics in
  the species pool
• Negative due to constraints on total richness at
  very small scales

20
Next real data

• To test whether observed patterns of native and
  exotic species richness are different from
  pattern generated by random assembly (the null
  expectation)
• Randomize native and exotic species labels in the
  species pool

21
Permutation tests for observational data

• Species pool
• Sp A
• Sp B
• Sp C
• Sp D

Nativity Label Native Exotic Native Native

• Calculate correlation coefficient (r) or slope
  (s)

• Repeat 500x, compare null distribution to real
  value

22
2
2
2
2
100 m
1 m
0.1 m
0.01 m
3
3
4
40
3
30
2
2
2
Exotic species richness
20
1
1
1
10
0
0
0
0
0
2
4
6
8
10
0
1
2
3
4
5
0
5
10
15
20
0
20
40
60
80
100
Native species richness
23
(No Transcript)
24
Coastal plant communities, 24 sites at 500 m2
14
12
10
Site exotic richness
8
6
4
4
6
8
10
12
14
16
Site native richness
25
Coastal plant communities, 24 sites at 500 m2
26
Conclusions

• The native-exotic richness relationship is
  scale-dependent, BUT, this is the null
  expectation
• With our null model competition for space
  among individuals, not species
• It is important to consider the null expectation
  when evaluating mechanistic explanations for
  patterns in data

27
Acknowledgements
Advising and Discussion Bob Peet, Peter White,
Jim McNair, UNC Plant Ecology Lab Funding
National Science Foundation, UNC Graduate School,
UNC Department of Biology, UNC Ecology
Curriculum, Sigma Xi, The Nature Conservancy,
USDA National Forest Service, Patrick Center for
Environmental Research Field crews