Drought, resilience and
response diversity of a semiarid annual plant community
Bertrand Boeken1 and Moshe Shachak2
1 Wyler
Dept. of Dryland Agriculture and
2 Mitrani
Dept. of Desert Ecology,
Blaustein Institute for Desert Research, Ben-Gurion Univ. of the Negev, Sde Boker Israel
1
Drought, resilience and
response diversity of a semiarid annual plant community
•
•
•
•
•
•
•
General question
Hypothesis
Research site
Two-year drought
Changes in annual plant density
Response groups
Specific mechanisms
Relationship between diversity
and ecosystem function
What aspect
of diversity ?
• Species
• Functional groups
• Attributes
?
Which ecosystem
Function ?
• Kinds of functions
- production
- decomposition
- nutrient cycling
- etc.
• Magnitude
• Stability, Resilience
3
Proposed hypothesis
Stress
Species’
response
diversity
Ecosystem
resilience
Ensures
• Self-organization
• with a variety of
responses to the same
stress
(Elmqvist et al. 2003)
4
• Capacity to learn and
adapt
Ecosystem function
• Species within a
functional group
High RD
• Recovery
• No state change
- the same controls on
structure and function
(Holling 1973, Carpenter et al.
2001)
low RD
Stress
Testing the hypothesis
Required
In semi-arid shrubland
1. Functional group with
many species
1. Species-rich annual plant
communities
2. Factor causing stress
2. Rare two-year drought event
3. A function with possible 3. Plant density and biomass
resilience
production
5
4. Variation in response
diversity
4. Two patch types with different
species
5. Long-term data
5. LTER site with annual
measurements
Sayeret Shaked Park ILTER
Northern Negev Desert, Israel
• Rainfall between November and
March
• Long-term mean annual rainfall ca.
200 mm/yr
• Mean min. winter temp. 6°C
• Mean max. summer temp. 34°C
300
Ofakim
200
Beer Sheva
150
100
Sde Boker
75
Mitzpe Ramon
50mm
N
0
40km
Eilat
6
Two-year drought
Drought years
Annual rainfall < 50%
of long-term average
Once in 8 yr
Rainfall (mm)
Annual rainfall 1998-2001
250
200
Long-term average
Ten-year
average
150
100
50
0
1998 1999 2000 2001
7
Two years in a row
Once in > 60 yr
Sampling design
Field sampling
Hill top
A
A north-facing slope (ca. 12%)
Four elevations, 16 plots (10 m x 8 m)
B
Herbaceous vegetation samples
(10 cm x 30 cm, N ≈ 98)
C
In 3 shrub and 3 crust patches per plot.
D
Wadi
Landscape patches
Scattered shrubs and
Crusted intershrub matrix
Measurements per sample
Counts of all individuals per species
Non-destructive sampling
Density as proxy variable for Biomass Production
8
No. of plants
Annual plant density in semiarid shrubland
120
80
40
0
1998 1999 2000 2001
Crust
Resilience
9
1998 1999 2000 2001
Shrub
Patch type
•
No state change - plant density and biomass maintained;
and no changes in control - still by annuals.
•
Differences between crust and shrub patches - crusts
don’t recover completely one year after the drought.
Mean no. of plants
Response groups
80 Crust
70
80 Shrub
70
60
60
50
50
40
40
30
30
20
20
10
10
0
1998 1999 2000 2001
Decrease during drought
10
0
Pre-drought
dominants
Subordinates
1998 1999 2000 2001
Recovery after drought
• Stronger in dominants
• Subordinates increase
• On the crust delayed
• In shrub patches also the
dominants
Pre-drought dominants
Mean no. of plants
50
Bromus
fasciculatus
40
30
Crust
Shrub
20
10
0
50
Mean no. of plants
Stipa
capensis
- Due to seed limitation?
‘98 ‘99 ‘00 ‘01
Anagallis
arvensis
‘98 ‘99 ‘00 ‘01
Silene
colorata
40
30
• In shrub patches,
dominant dicots suffer
too, but recover faster.
- Due to site limitation?
20
10
0
‘98 ‘99 ‘00 ‘01
11
• Dominant grasses
suffer during drought,
and recover slowly.
‘98 ‘99 ‘00 ‘01
Subordinate species
Mean no. of plants
50
40
30
20
Crust
Shrub
10
0
50
Mean no. of plants
Reboudia
pinnata
40
‘98 ‘99 ‘00 ‘01
Ammochloa
palaestina
Filago
contracta
30
• Many subordinate species
contribute to post-drought
recovery in both patch types.
- Due to release of site or resource
competition by dominants?
20
10
0
‘98 ‘99 ‘00 ‘01
12
• Subordinate species suffer little,
increase during and after the
drought.
‘98 ‘99 ‘00 ‘01
Response diversity
3 response groups
1. Dominant grasses S. capensis and B.
fasciculatus
• decrease during drought and
• recover slowly
2. Dominant dicots A. arvensis and S.
colorata
• decrease during drought and
• recover faster
• only dominant in shrub patches
3. Subordinates
• increase during drought
• major contribution to
recovery
• especially R. pinnata
13
Dominants
Mechanisms
S. capensis and B. fasciculatus in both patch types
! Reduced seed production
! Seed-limited recruitment (Boeken and Shachak
1998)
! Have no seed dormancy (Boeken et al. 2004)
A. arvensis and S. colorata in shrub patches
• Limited germination sites
• Recruitment from a dormant seed bank
Subordinates in both patch types
• Dormant seed bank
R. pinnata:
! Better germination in light
(Zaady et al. 1997) on exposed
surface with less litter.
14
Conclusions
Shrub patches: High RD
Resilience of ecosystem function
(perhaps if drought more
prolonged?)
• Re-organization of species ranks
• Shrub patches recover better than
crusted intershrub matrix
Primary production
• No state shift
Crust: low RD
System
collapse
System
recovery
Drought stress
Response diversity
• Shrub patches - 3 groups
15
• Crusted matrix - 2 groups
- Dominant grasses
- Dominant grasses
- Dominant dicots
- No dominant dicots
- Many subordinates
- Fewer subordinates
Thank you
16