How interaction networks help to understand ecosystem functioning and stability
ecosystem functioning
0 11 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 01 0 0 1 0 0 0 1 0 1 0 1 1 1 1 1 0 1
global change
H2'=0.26
72210011 20001100 10000000
Nico Blüthgen AG "Ecological Networks", Biology, Technische Universität Darmstadt
SPECIES INTERACTION NETWORKS
Functional relationships consumer – resource pollination dispersal ...
SPECIES INTERACTION NETWORKS
generalized & redundant
specialized & complementary
Robust: insurance hypothesis
Fragile to co-extinctions
Stabilizing portfolio effect
Higher fluctuations
NETWORKS: PROBLEMS
food webs
interaction networks
loops
interaction strength (link weight) based on species list, abundance and body mass
based on number of observed interactions, e.g. visits by pollinator individuals sampling limits observations per species / per network!
NETWORKS: PROBLEMS
interaction strength or ‘dependence’ = 1.0
These Metrics are directly affected by variation in total number of observations (sampling, abundance) per species / per network also: nestedness, connectance…
'generalist'
0.2
0.7
0.1
Number of observations
Number of links
'specialist'
Interaction strength
‘Classical’ specialisation metrics in networks Number of links (plant species)
Number of observations Blüthgen (2010) Basic Appl Bascompte Ecol * ‘Dependence’ sensu et al. (2006) Science
NETWORKS: PROBLEMS
Pollinator species
673 154 67 110 1 5 8
5
4 9 8
4
link weight (here: visits)
6
1
7 1 4 6 1 4 3 4 4 3
2 3
2 1 1 1 1 1
∑
1 790 208 72 20 15
673 221 116 16 15 14 1 9 9 8 7 7 3 1 7 6 5 4 4 4 4 3 3 3 2 2 2 1 1 1 1 1 1 15 5 4 1 1130
Frequency
∑
Plant species
Blüthgen et al (2006) BMC Ecol
NETWORKS: SPECIALIZATION METRIC Null model Real network H2’=0.85 811 6 15 2243 87 54 15 12 1 32 4 22 2 2 34 1 1 1 2 1
9
1
9 8
1 1
accounts1 for variation in 1 1 total observations per species
∑
790
208
72
20
15
1 1
1
Pollinator species
1
14 1
7 7 7 6 5 4 4 4 4 3 3 2 2 1 1
5
(marginal totals fixed)
673
1 1130
67
110 1
221 116 16
6 15 5
5
4
14
8
1 9 8
8 7
1
2
4 6
3
3
1
7 7 6
1 4
5 4
1
3
4
4 4
4 4
3 3 2 2 1 1 1 1 1
∑
9 9
4
1 1
4
673 154
7
1 1 15
∑
Plant species
11 673 221 116 16
(marginal totals fixed)
Pollinator species
471 469 127 42 11 8 460 126 129 46 16 10 484 113 38 42 11 10 12 481 120 38 13 78 161 37 13 555 13 2 169 41 31 154 137 56 18 15 42 154 37 16 42 81 20 323 11 22 78 24 896 21 77 26 96 8482 15 21 9 4 1 1 12 1010 56 11 2 3 121 912 51 43 1 9 4 87 46 1 22 1 79 14 48 4 11 1 11 86 2 5 2 3 11 1 66 22 12 1 4 11 11 1 55 11 6 6 1 12 57 221 4 1 1 6 1 52 12 1 3 34 1 23 2 52 121 3 3 1 34 122 1 3 32 11 34 11 3 3 24 21 3 34 111 3 3 1 24 111 1 3 1 1 12 21 1 11 1 1 32 21 211 1 2 1 11 11 1 11 1 11 1 1 1 1 1 1 11 1 1 1 1 1
1 790
208
72
20
15
15
5
4
3 3 2 2 1 1 1 1 1 1 1 1130
Frequency
∑
Plant species
Blüthgen et al (2006) BMC Ecol
NETWORKS: SPECIALIZATION METRIC Null model Real network 124 41 21 3 3 2 2 1 1 1 1 1 1 1 1 1 1 1 1
790
208
43 12 7 1 1 1 1 1 1 1 1 1 1
13 4 2 1
9 4 2
9 4 2
3 1 1
3 1
1
673 221 116 16 14 9 9 8
Pollinator species
Pollinator species
∑
468 154 81 11 10 6 6 6 5 5 5 4 3 3 3 3 3 2 2 2 2 1 1 1 1 1 1
673 154
15
5
4
15 5 8
5 4 Specialist 1 9 di’=0.9 8
1
2
7 6
4 6
3
5 4
4
4
1
1 1130
221 116 16
6
7 3
1
Opportunist 3 4 di’=0.0 4
di‘ for each species i 2
2
1 1 1 1
208
8 7
5 4 4 4 4 3 3 2 2 1 1 1 1
1
∑
9 9
7
1 4
3
1 790
14
7 6
3
1 1 1 1 15
67
7 7
3 3 2 2 1 1
20
673
110 1
4 4
72
∑
Plant species
72
20
15
15
5
4
Frequency
∑
Plant species
1 1 1 1130
Blüthgen et al (2006) BMC Ecol
NETWORKS: SPECIALIZATION METRIC Complementary specialization Pollination web (H2) Most specialised (H2min)
Most generalised (H2max) 468 154 81 11 10 6 6 6 5 5 5 4 3 3 3 3 3 2 2 2 2 1 1 1 1 1 1 790
124 41 21 3 3 2 2 1 1 1 1 1 1 1 1 1 1 1 1
43 12 7 1 1 1 1 1 1 1 1 1 1
13 9 9 3 3 1 673 4 4 4 1 1 221 2 2 2 1 116 1 16 14 9 9 8 7 7 7 6 5 4 4 4 4 3 3 2 2 1 1 1 1 1 1 208 72 20 15 15 5 4 1 1130
673 154 67
r
c
H 2 max − H 2 H 2 max − H 2 min
5
4 9 8
i =1 j =1
H2'=
6 5 8
H 2 = −∑∑ ( pij ⋅ ln pij ) 4 1
7 1 4 6 1 4 3 4 4 3
2 3
2 1 1 1 1 1 1 790 208 72 20 15
redundant
0.0
110 1
Shannon entropy
673 221 116 15 16 14 1 9 9 8 7 3 1 7 7 6 5 4 4 4 4 3 3 3 2 2 2 1 1 1 1 1 1 15 5 4 1 1130
673 208 13 116 16 14 9 9 8 7 7 6 5 4 4
2 1
1 1 1
790 208 72 20 15
673 221 116 16 14 9 9 8 7 7 7 7 6 5 4 4 4 4 4 4 3 3 3 3 2 2 1 1 1 1 1 1 1 1 1 1 15 5 4 1 1130
complementary
0.5
H 2’
0.85
1.0
Blüthgen et al (2006) BMC Ecol
NETWORKS: SPECIALIZATION PATTERNS
Asymmetric specialization? (“nestedness”)
... plus dozens of follow-up studies!
NETWORKS: SPECIALIZATION PATTERNS
y c n e Asymmetric specialization? (“nestedness”) u freq Pollinator species
Plant species
∑
468 154 81 11 10 6 6 6 5 5 5 4 3 3 3 3 3 2 2 2 2 1 1 1 1 1 1
124 41 21 3 3 2 2 1 1 1 1 1 1 1 1 1 1 1 1
790
208
43 12 7 1 1 1 1 1 1 1 1 1 1
13 4 2 1
9 4 2
9 4 2
3 1 1
3 1
1
673 221 116 16 14 9 9 8
7 Nestedness: 7 7 • observations limited 6 • common and rare54 species 4 • no extreme specialization 4 4 3 3 2 2 1 1 1 1 1 1 72
20
15
15
5
4
1 1130
„Asymmetries“ are a direct consequence of variation in observation totals per species (null model, arithmetric proof) Blüthgen (2010) Basic Appl Ecol
NETWORKS: SPECIALIZATION PATTERNS
Number of links
Are rare species more specialized?
Number of observations
Plants: r = 0.27*
no!
Number of observations *Meta-Analysis of 20 networks (global)
Specialization (dj’)
Specialization (di’)
Pollinators:
Specialization is variable, independent of abundance r = –0.20*
yes!
Number of observations Blüthgen et al (2007) Curr Biol
NETWORKS: SPECIALIZATION PATTERNS redundant generalised
complementary specialised
n = 162 pollinator networks (Alb/Hainich/Schorfheide 2008) x ± SD Weiner et al (submitted) n = 40 pollinator networks (Alb 2007)
Weiner et al (2010) Basic Appl Ecol
n = 27 pollinator networks (Würzburg 2006)
Fründ et al (2010) Oikos
0.0
H 2’ 0.5 Complementary specialization
1.0
NETWORKS: SPECIALIZATION PATTERNS
N = 80 regions (282 networks) Schleuning, Fründ et al. (2012) Curr Biol
NETWORKS: SPECIALIZATION PATTERNS
Schleuning, Fründ et al. (2012) Curr Biol
NETWORKS: SPECIALIZATION PATTERNS Complementary specialization (n = 21)
***
Pollinator–Plant (n = 8)
WhichFrugivore–Plant traits structure these networks?
(n = 14)
Ant–Myrmecophyte (n = 8)
Ant–EFN
***
Herbivore–Plant plant defenses!
0.0
0.5
H 2’
1.0
Blüthgen et al (2006) Insectes Soc Blüthgen et al (2006) J Trop Ecol Blüthgen et al (2007) Curr Biol
NETWORKS: TRAITS PATTERN
Which traits structure each link? Real network H2’=0.82
Summe
468 154 81 11 10 6 6 6 5 5 5 4 3 3 3 3 3 2 2 2 2 1 1 1 1 1 1
124 41 21 3 3 2 2 1 1 1 1 1 1 1 1 1 1 1 1
790
208
43 12 7 1 1 1 1 1 1 1 1 1 1
13 4 2 1
9 4 2
9 4 2
3 1 1
Plant species
Summe
3 1
1
673 221 116 16 14 9 9 8
Pollinator species
Pollinator species
Plant species
20
15
15
5
4
673
673 154
67
110 1
221 116 16
6 15 5
5
4
14
8
1 9
7 7
Hot 8links 7 1Cold links 2 3
7 6
4 6
9 9 8 7
1
7
3
7 6
5 4
4
1 4
5 4
4
1
3
4
4 4
4 4
4 4
3 3 2 2 1 1
3
1 1
1 1
3 2 2 1 1
1 1 72
Summe
1 1130
Neutral
1 Summe
1 790
208
72
20
15
15
5
4
3 3 2 2 1 1 1 1
1 1 1 1130
Frequency
H2’=0.0 Null model
NETWORKS: TRAITS PATTERN
Olfactory signals structure flower visitor networks?
Outdoor olfactometer
Control
Flower odor
Junker et al. (2010) J Anim Ecol
NETWORKS: TRAITS PATTERN
Olfactory signals structure flower visitor networks! r2 = 0.16*** + morphology + color + ressources...
cold link
Flower odor
less than expected
Control
hot link
more than expected
interactions
repelled
signal response
attracted
Junker et al. (2010) J Anim Ecol
NETWORKS: TRAITS PATTERN CONSEQUENCES Floral scents have evolved as defensive trait (against ants)? Metrosideros polymorpha endemic
Hawai'i
NETWORKS: TRAITS PATTERN CONSEQUENCES
Hawaiian flowers are invaded by ants hot link
0.2
a
a
b
n = 17
n = 20
n = 18
0.1
0
-0.1
cold link
more than expected
interactions with ants
less than expected
Metrosideros polymorpha endemic
-0.2
F = 3.7* -0.3
Endemic Lantana camara: invasive
Native
Introduced
Plant species Junker et al. (2011) Ecol Monographs
NETWORKS: PATTERNS CONSEQUENCES 3 regions 3 x 50 meadows/pastures
162 single-day networks: (each 600 m2, 6h) 119 grasslands 25401 visits (individuals) of 741 pollinator spp. on 166 plant spp. ommunity stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability
lity for total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces nd (f) has no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was anisms (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that phic control of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of d enhances portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant lant community biomass (Tilman et al. 2006b).
of the unexplained al. 2001). otential to dampen ponses of predator with responses of ty of predator and = 0.09, P = 0.001), rsity (McCann et al. undance responses
variability in total herbivore abundances and diversity. Plant diversity thus provides a key ecosystem service by reducing insect outbreak potential defined both in terms of herbivore population and community abundance and herbivore community variability over time. This service can be enhanced through landscape management for biodiversity, such as in conservation or restoration of natural areas in agricultural landscapes and diverse plantings for biofuels production (Tscharntke et al. 2005; Losey & Vaughan 2006; Landis et al. 2008; Isaacs et al. 2009). More broadly, our work suggests that biodiversity conservation or restoration at the producer trophic level contributes to the maintenance of diversity, function, and stability of entire foodwebs.
Land use intensity
60 40
30
20
20
60
100
140
0.0
Fertilization
2.0
3.0
0 200
600
1000
Grazing
[livestock units ⋅ days ⋅ ha-1 ⋅ year-1]
[year ]
Fi Mi C + + i Fmean M mean Cmean
0
10
20
30
40
1.0
Mowing -1
[kg nitrogen ⋅ ha-1 ⋅ year-1 ]
Li =
0
10 0
0 20
Number of plots
80
50 40
80 60 40 20 0
Number of plots
60
LAND USE
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Land use intensity (LUI) Li ' = Li Blüthgen et al. (2012) Basic Appl Ecol
LAND USE BIODIVERSITY
Land use intensity Blüthgen et al. (2012) Basic Appl Ecol
LAND USE: WINNERS AND LOSERS 25
Zygaena purpuralis
r
Lose r
0
10
5
20
10
15
40 30
e Winn
0.5
1.0
1.5
2.0
Land use intensity
2.5
0.5
3.0
1.0
1.5
2.5
3.0
rS = –0.35*** Cryptocephalus aureolus
10
15
3 2
Lasioglossum albipes
2.0
Land use intensity
rS = –0.16*
Lose r
0
5
1
Lose r
0
Number of individuals
rS = –0.31***
20
50
Episyrphus balteatus
0
Number of individuals
60
rS = 0.09ns
0.5
1.0
1.5
2.0
Land use intensity
2.5
3.0
0.5
1.0
1.5
2.0
Land use intensity
2.5
3.0
n=162
networks
NETWORKS: PATTERNS CONSEQUENCES
-0.2
-0.1
0.0
0.1
0.2
r = –0.13, p = 0.003
-0.3
bees butterflies beetles hoverflies other dipterans other hymenopterans
-0.4
Loser
Land use effect [rS*]
Winner
0.3
Can networks predict biodiversity declines?
0.0
*) rS: land use ~ abundance
0.2
0.4
0.6
Specialization [di’]
0.8
1.0
n = 741 visitor species
Specialized consumers suffer disproportionally from land use intensification
Weiner et al. submitted
NETWORKS: PATTERNS CONSEQUENCES 0.3 0.2
r = 0.51, p < 0.0001
-0.2
-0.1
0.0
0.1
Winner
-0.3
bees butterflies beetles hoverflies other dipterans other hymenopterans
-0.4
Loser
Land use effect on consumer [rS*]
Can networks predict biodiversity declines?
-0.4
-0.3
Loser
-0.2
-0.1
0.0
0.1
0.2
0.3
Winner
Land use effect on plant species visited [rS*] *) r : land use ~ abundance = 767 visitor species The Sfate of a plant determines the fate of its consumers (→n interaction strength)
The reverse effects of pollinators on plants is much weaker (r Weiner = 0.33***) et al. submitted
NETWORKS: PATTERNS CONSEQUENCES
Redundancy increases stability? Environmental complementarity
2
2
2
Functionally 'redundant’ pollinator species
environmental conditions or disturbance regimes
Blüthgen & Klein (2011) Basic Appl Ecol
NETWORKS: PATTERNS CONSEQUENCES (f) Redundancy = "effective" richness eH of each species' partners
(e)
6 2
4
Redu
y
8
1.0
1.5
2.0
2.5
rS= –0.76, p = 0.02, n = 12 plots
3.0
Land use intensity (LUI) r = 0.07, p = 0.006
6
0.5
40
ndanc
Lepidoptera: Plebeius argus
0
Redundancy Redundancy [eH]
8
Spearman‘s rs = –0.06, p < 0.0001
0
4 2
ndancy u d e R variability in total herbivore abundances and diversity. Plant diversity thus provides a 0
namics beyond the plot scale could help account for some of the unexplained in arthropod community dynamics over time (Haddad et al. 2001). fects of plant diversity on higher trophic levels have the potential to dampen asing distance up the food chain. Yet, we found that responses of predator itoid species richness to plant diversity were consistent with responses of 0.5 1.0 1.5 2.0 2.5 3.0 species richness. As plant diversity increased, the stability of predator and species richness increased by 33% (Fig. 2a; n = intensity 163, r2 = 0.09, (LUI) P = 0.001), Land use usediversity intensity y as a result of higher habitat structure and Land herbivore (McCann et al. ddad et al. 2009). We found that predator and parasitoid abundance responses e variable. For the seven most abundant predator and parasitoid species,
Redundancy
30 20 10
Redundancy
Redundancy [eH]
Land use Effects of plant species richness on arthropod population andintensity community stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability s richness; (b) reduces predator and parasitoid community stability for total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces stability of most herbivore specialists, including Aphis spp.; and (f) has no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was as cv)1 at the population and community levels. Potential mechanisms (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that rsity (1) increases predator abundances and potential for trophic control of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of ng (Haddad et al. 2009); (3, 4) reduces species covariances and enhances portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant y biomass (Tilman et al. 2001), and increases the stability of plant community biomass (Tilman et al. 2006b).
key ecosystem service by reducing insect outbreak potential defined both in terms of herbivore population and herbivore3.0 community variability 0.5 1.0community 1.5 abundance 2.0 and2.5 over time. This service can be enhanced through landscape management for biodiversity, such as in conservation or intensity restoration of natural areas in agricultural Land use landscapes and diverse plantings for biofuels production (Tscharntke et al. 2005; Losey & Vaughan 2006; Landis et al. 2008; Isaacs et al. 2009). More broadly, our work suggests that biodiversity conservation or restoration at the producer trophic level contributes to the maintenance of diversity, function, and stability of entire foodwebs.
NETWORKS: PATTERNS CONSEQUENCES
Response diversity Environmental complementarity How different are their response traits? 2
2
2
Functionally 'redundant’ pollinator species performance measures morphological traits physiological traits life-history traits phenology etc.
environmental conditions or disturbance regimes
phylogeny
Elmquist et al. (2003) Frontiers Ecol Environm, Laliberte et al. (2010) Ecol Lett
2.0 0.5
1.0
1.5
... of redundant pollinators rs = –0.08, p = 0.002
0.0
0.5
1.0
Redundancy
1.5
2.0
... of redundant plants rs = –0.05, p < 0.001
0.0
Phylgenetic distance Mean phylogenetic distance*
NETWORKS: PATTERNS CONSEQUENCES
0.5
1.0
1.5
2.0
2.5
3.0
Land use intensity Land use intensity (LUI)
0.5
1.0
1.5
2.0
2.5
3.0
Land use Landintensity use intensity (LUI)
*Grafen (1989) method molecular tree (plants) classification tree (pollinators)
NETWORKS: PATTERNS CONSEQUENCES (d)
's viewpoint:
Redu
(f)
ndanc
Phylo genet ic
's viewpoint:
y c n a und
y
divers ity
Red Phylo genet
ic div e
rsity
Land use
Biotic homogenization Vulnerability to climate change?
d population and community stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability d community stability for total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces ding Aphis spp.; and (f) has no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was els. Potential mechanisms (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that potential for trophic control of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of es covariances and enhances portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant s the stability of plant community biomass (Tilman et al. 2006b).
NETWORKS: PATTERNS CONSEQUENCES
Redundancy increases stability!!! Spearman‘s rs = 0.41, p < 0.001
2
2
Functionally 'redundant’ pollinator species
10.0 5.0
●
0.5
2
●
● ●● ●
1.0
r = 0.42
●
2.0
-1 Stability Stability[CV ]
20.0
●
● ● ● ●
●
●
● ● ● ● ● ●
●
● ● ●
●●● ● ● ● ●● ● ● ● ● ●● ●
● ● ●
● 5
10
15
Shannon diversity (e^H)H Redundancy [e ]
5 years
Vazquez, Chacoff, Blüthgen (unpubl.)
Main contributors to unpublished results shown in this talk:
Oliver Mitesser Christiane Weiner Michael Werner Diego Vazquez Natacha Chacoff
Thank you!
NETWORKS: PROBLEMS
Some clear ecological concepts ... e.g. observation hours indiv.-1 species-1
• Abundance
e.g. individuals time-1 area-1
• Biodiversity
e.g. Shannon Diversity
• Niche breadth / overlap
Activity
• Sampling
Different ressources or conditions
NETWORKS: PROBLEMS
Some clear ecological concepts ... Network metrics:
• Sampling
• Nestedness
• Abundance
• Connectance
• Biodiversity
• Generality
• Niche breadth / overlap
• Interaction diversity
... are blurred in most network metrics! Blüthgen (2010) Basic Appl Ecol
ROLE OF BIODIVERSITY 1. Biodiversity Ecosystem Functioning Species complementarity 2. Biodiversity Functional Stability
Functioning
Species redundancy redundant
Biodiversity
ROLE OF BIODIVERSITY 1. Biodiversity Ecosystem Functioning Species complementarity 2. Biodiversity Functional Stability
Functioning
Species redundancy
redundant
Biodiversity
ROLE OF BIODIVERSITY BD Ecosystem Functioning BD (Functional) Stability Stability = CV–1
39 Tilman
et al. (unpubl.)
www.cedarcreek.umn.edu
NETWORKS: PROBLEMS
Asymmetric specialization? (“nestedness”)
NETWORKS: PROBLEMS
Asymmetric specialization? ... where to go?
... specialise on happy flowers with the greatest pollinator spectrum?
.. reliability!
... or on flowers with few pollinator species?
... avoiding competition!
Benadi et al. (2012) Am Nat
3.0 1.5
2.0
2.5
se u d lan
1.0
Flower diversity [H]*
Diversity of flower visitors [H]
3.5
Diversity of flower visitors [H]
Biodiversity decline
0.0
0.5
1.0
1.5
2.0
2.5
Flower diversity [H]*
H2’ = 0.51 ± 0.11 SD r = 0.49**, n = 27 networks (Würzburg) Fründ et al. (2010) Oikos
H2’ = 0.62 ± 0.16 SD rS = 0.41***, n = 162 networks (Schwäb. Alb, Hainich, Schorfheide) *based on flower area
Weiner et al. (submitted)
Mutualistic networks
+
Plant-Pollinator networks
Food webs (Trophic networks) Plant-Herbivore networks
+
+
–
Plant-Frugivore networks
Predator-Prey networks
Plant-Ant networks
Parasite-Host networks
REDUNDANCY: RESPONSE DIVERSITY Defining response diversity Functionally 'redundant' animal species sp1 sp2 sp3
How different are their response traits?
...
low response diversity Rk sp1 sp2 sp3 r1 stability r2 r3 munity stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community
for total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces f) has no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was ms (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that control of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of hances portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant 1 2 community biomass (Tilman et al. 2006b).
he unexplained 2001). ntial to dampen ses of predator h responses of of predator and 09, P = 0.001), (McCann et al. dance responses
high response diversity Rk sp1 sp2 sp3 r r r3
Response trait (ri ) variability in total herbivore abundances and diversity. Plant diversity thus provides a key ecosystem service by reducing insect outbreak potential defined both in terms of e.g. surface/volume ratio herbivore population and community abundance and herbivore community variability pubescence density over time. This service can be enhanced through landscape management for biodiversity, such as in conservation or restoration of natural areas in agricultural larval habitat landscapes and diverse plantings for biofuels production (Tscharntke et al. 2005; Losey overwintering stage... & Vaughan 2006; Landis et al. 2008; Isaacs et al. 2009). More broadly, our work suggests that biodiversity conservation or restoration at the producer trophic level contributes to the maintenance of diversity, function, and stability of entire foodwebs.
Networks: patterns processes consequences
Can networks predict biodiversity declines? Number of links
Do "generalists" require (or profit from) a variety of resources / plant species? Temporal / conditional complementarity
Number of observations
Observed / realized niche Fundamental niche?
Overestimate co-extinction?
environmental conditions daytime or season Nutritional complementarity sugar
protein
or toxin dilution etc.
Underestimate co-extinction? Blüthgen (2010) Basic Appl Ecol
Blüthgen & Klein (2011) Basic Appl Ecol
higher redundancy
REDUNDANCY: DRIVERS 1
2
1
1
1
1
1
1
1
lower density
lower total diversity
higher specialization high H2’
abundance 2 detection probability
2
sampling?
2
higher density
specialization for given density and diversity low H2’ higher generalization
higher total diversity 3
3
3
3
3
3
3
3
3
REDUNDANCY: DRIVERS
1) Abundance?
y
0 1 2 3 4 5 6
y
log(# of individuals)
ndanc
Redundancy
Redu
per pollinator species
Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability ance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces ve effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was s) show the results of this and prior studies that inform arthropod responses in this study, including that bivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of o effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant omass (Tilman et al. 2006b). s
ACKNOWLEDGEMENTS
1.0
1.5
2.0
2.5
3.0
0 1 2 3 4 5 6 7
log(# of individuals)
0
200
variability in total herbivore abundances and diversity. Plant diversity thus provides a key ecosystem service by reducing insect outbreak potential defined both in terms of herbivore population and community abundance and herbivore community variability over time. This service can be enhanced through landscape management for biodiversity, such as in conservation or restoration of natural areas in agricultural landscapes and diverse plantings for biofuels production (Tscharntke et al. 2005; Losey & Vaughan 2006; Landis et al. 2008; Isaacs et al. 2009). More broadly, our work suggests that biodiversity conservation or restoration at the producer trophic level contributes to 0.5 1.0function, 1.5 and2.0 the maintenance of diversity, stability2.5 of entire3.0 foodwebs.
LandLand use use intensity intensity(LUI)
0.5
per plant species Land use intensity rs = 0.07, p = 0.01
Redundancy
r = 0.01, p = 0.85 per site
600
Redundancy
n r f d , l. s , h e
# of individuals (visits)
d
c n a d n u Red
rs = 0.02, p = 0.21
0.5
1.0
1.5
2.0
2.5
3.0
Land use intensity Land use intensity (LUI)
REDUNDANCY: DRIVERS
8 10 6 4 2 0
0.5
1.5
2.0
2.5
3.0
40
Land use intensity rs = 0.14, p = 0.08
0
nity stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces as no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability0.5 was (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that ntrol of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of ces portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant mmunity biomass (Tilman et al. 2006b).
Land use intensity
1.0
30
y
Redundancy
c n a d n u Red
20
y
rs = –0.21, p = 0.008
10
ndanc
Total pollinator diversity [eH]
Redu
Redundancy Total flower diversity [eH]
2) Diversity?
1.0
1.5
2.0
2.5
3.0
intensity (LUI) Land Land use use intensity
REDUNDANCY: DRIVERS
0.8
1.0
rs = 0.11, p = 0.22
0.6
Alb rs = 0.32, p = 0.03 Hainich rs = 0.20, p = 0.18 Schorfh. rs = -0.22, p = 0.19
0.4
Complemantary specialisation [H2‘]
3) Specialization?
0.5
1.0
1.5
2.0
2.5
Land use intensity (LUI)
3.0
REDUNDANCY: DRIVERS
Drivers of redundancy loss through land use: Density
(0)
Diversity
✔
Specialization
(0✔)
✔ ✔ ✔
ommunity stability. Higher plant diversity (a) increases predator and parasitoid (abbreviated "Predator!) community stability lity for total abundance; (c,d) increases herbivore community stability for species richness and total abundance; (e) reduces nd (f) has no negative effect on the stability of most herbivore generalists, including Lygus lineolaris. Temporal stability was anisms (right panels) show the results of this and prior studies that inform arthropod responses in this study, including that phic control of herbivores (Haddad et al. 2009); (2) reduces herbivore abundances and thus eliminates the possibility of d enhances portfolio effects (this study); and (5) decreases plant population stability (Tilman et al. 2006b), increases plant lant community biomass (Tilman et al. 2006b).
of the unexplained al. 2001). otential to dampen ponses of predator
variability in total herbivore abundances and diversity. Plant diversity thus provides a key ecosystem service by reducing insect outbreak potential defined both in terms of herbivore population and community abundance and herbivore community variability over time. This service can be enhanced through landscape management for
Białowieza Forest: old growth versus logged
Albrecht et al. submitted