Transcript ppt

Multiple-species interactions
Left: Image from Wikimedia Commons of one of the earliest known depictions of a food web, by Victor
Summerhayes & Charles Elton (1923) for Bear Island, Norway
Right: Provenance of “A simplified food web for Northwest Atlantic” unknown
Food Webs
Trophic (energy & nutrition) relationships among organisms
Nodes
Taxonomic or functional categories
Links
Flow of material (including
energy-rich molecules)
Paine, R. T. (1966) – Food webs are
the “ecologically flexible scaffolding
around which communities are
assembled and structured”
Provenance of image unknown
Food Webs
Elton (1927) observed that predators tend to be larger & less numerous
than their prey – “pyramid of numbers” (a.k.a. “Eltonian pyramid”)
Elton’s hypothesis: Predators must be larger than prey to subdue them
Pyramid could
represent numbers,
biomass, energy
consumed per year,
etc.
Image from http://mrskingsbioweb.com/ecology.html
Food Webs
Lindeman (1942) introduced the “energy-efficiency hypothesis” – the
fraction of energy entering one trophic level that passes to the next
higher level is low (~ 5 - 15%)
The first and second laws of thermodynamics
predict inefficiency:
1st Law = Conservation of Energy
2nd Law = Energy transformations result in
an increase in entropy, i.e., only a
fraction of the energy captured by
one trophic level is available to do
work in the next
Inverted pyramids of biomass can occur (e.g., whales, krill,
phytoplankton in southern oceans), but only when productivity and
turnover of producers is extremely high
Food Webs
Levin, S. A. (1992) – “Is a taxonomic subdivision most appropriate… would a
functional one serve better? Should subdivision… consider different
demographic classes, be partitioned according to genotype, etc.?”
2 Consumers
Trophic levels
within a simple
food chain;
donor levels
supply energy
or nutrients to
recipient levels
1 Consumers
“Green” or living
food web
1 Producers
1 Consumers
2 Consumers
“Brown” or detrital
food web
Food Webs
Web jargon:
Connectance (c): Number of links (L) or connections between species
(S) or nodes – expressed as a proportion of maximum connectance:
c = L / [S(S-1)/2]
Maximum connectance = S(S-1)/2
Linkage density (L/S): Average number of trophic links per species
Compartmentation: Degree of isolation of subwebs – the number of
species that interact with any given pair of species versus those that
interact with only one member of the pair
Food Webs
Web jargon:
Omnivory: Feeding on more than one trophic level
4
Same-chain
omnivory
4
3
3
3’
2
2
2’
1
1
1’
Different-chain
omnivory
Food Webs
Web jargon:
Cycles & loops: Species have reciprocal feeding relationships
Cycle
E.g., wasps that prey
on spiders that in turn
catch wasps in their
webs
Loop
E.g., “rock-paper-scissors”
interactions among
plankton
(see Huisman refs.)
A
B
A
B
C
Food Webs
Modeling food webs:
Which food web configurations promote stable equilibria?
May (1973) and Pimm & Lawton (1977, 1978) used multispecies LotkaVolterra models to examine various configurations for stability
1
2
3
4
1
-
-
0
0
2
+
0
-
0
3
0
+
0
-
4
0
0
+
0
4
3
2
1
0 = no connection / no interaction
+ = positive effect; prey supplying
energy to predator
- = negative effect; predation
Values corresponded to
interaction strengths
Food Webs
Simulations generally examine the influence of small changes in predator
& prey populations away from equilibria
Two criteria for assessing stability:
Do populations return to equilibrium sizes?
How long does the system take to return to equilibrium?
The way in which the matrices are constructed (e.g., lengths of food
chains, connectedness, etc.) determines stability
Do real-world food webs yield repeated patterns?
If so, do the patterns have ecological significance?
Food Webs
Are ratios of species at different trophic levels constant across
communities?
Cohen (1978) reviewed published community webs – relatively high
consistency of predators to prey (4:3)
This may simply reflect greater lumping into functional groups for prey
than predators
How long are food chains?
As expected, relatively short; rarely more than 5 trophic levels
(Pimm & Lawton 1977; Pimm 1982)
Food Webs
How long are food chains?
Yodzis (1984) – meta-analysis of 34 published food webs (Briand 1983)
to examine the influence of energy efficiency on food-chain length
Invertebrate ectotherms vs. vertebrate ectotherms vs. vertebrate
endotherms at trophic level 2
Energy-conversion efficiency:
invert. ectotherms > vert. ecototherms > vert. endotherms
(invert. ectotherms are about an order of magnitude
more efficient than vert. endotherms)
Percent of chains supporting consumer(s):
23%
>
9%
>
6%
invert. ectotherms > vert. ecototherms > vert. endotherms
Food Webs
How long are food chains?
Jenkins et al. (1992) – direct test of the energy-efficiency hypothesis
If efficiency of energy transfer primarily determines food chain length, then
manipulating productivity should influence food chain length
Plastic buckets in an Australian forest to resemble water-filled tree-holes
with different amounts of litter to generate a productivity gradient
Natural tree-holes contain 4-level trophic chains:
litter -- mosquito larvae -- larvae of predatory midge -- tadpoles
Litter at 100% natural level (938 g/m2/yr), 10% natural level, 1% natural level
Well-replicated study tracked for 48 wk
Food Webs
How long are food chains?
Jenkins et al. (1992) – direct test of the energy-efficiency hypothesis
Decreased productivity resulted in decreased number of coexisting
species & decreased number of trophic levels & links
Figure from Jenkins et al. (1992)
Food Webs
Modeling suggested that cycles, loops, and omnivory would destabilize
food webs
Do cycles and loops occur in nature?
Is omnivory common?
Polis (1991) – a skeptic of food web theory – characterized desert food
webs in great detail
Two-species cycles and three-species loops occur, and are especially
common in communities characterized by size-dependent predation
Role reversals between predators and prey are not uncommon
Omnivory is quite common
Interaction Webs
The distribution of interaction strengths is very important for determining
modeling outcomes
How are interaction strengths distributed in nature?
Unlike the randomly defined interaction strengths of the earliest modeling
approaches, interaction strengths are not normally distributed; they are
heavily skewed toward weak interactions
“…weak interactions may be the glue that binds natural communities
together” (McCann, Hastings & Huxel 1998)
This shows that evaluating interaction strength (of combined direct &
indirect effects) and not merely trophic links is essential to
understanding population dynamics and stability within food webs
Direct & Indirect Effects
Dissecting exploitation competition reveals its indirect nature
H
-
H
-
-
+
+
P
Solid arrows indicate direct effects, dotted arrows indicate indirect effects
Redrawn from Menge (1995)
Direct & Indirect Effects
Dissecting the ant-acacia mutualism reveals its indirect components
ant
H
-
-
-
+
+
+
P
As consumers, ants have direct negative effects on acacias (eating Beltian
bodies, etc.), but indirect positive effects mediated through herbivores
Solid arrows indicate direct effects, dotted arrow indicates indirect effect
Direct & Indirect Effects
Apparent
Competition
Tri-trophic Interaction
or Trophic Cascade
C
H
-
+
P
+
+
+
-
P
+
H
-
+
P
Solid arrows indicate direct effects, dotted arrows indicate indirect effects
Redrawn from Menge (1995) & Morin (1999)
Direct & Indirect Effects
Indirect
Mutualism
Keystone
Predation
-
+
H
H
+
-
+
-
+
H
H
+
-
P
P
-
P
H
-
+
+
P
Habitat
Facilitation
- (e.g., inhibits
feeding)
P
Solid arrows indicate direct effects, dotted arrows indicate indirect effects
Redrawn from Menge (1995) – found 83 distinct types of indirect interactions in 23 communities
Direct & Indirect Effects
Interaction chain indirect effect
– results from “linked direct interactions”
(e.g., bird predators enhance barnacle
abundance b/c they consume limpets that
dislodge & sometimes consume barnacles);
relatively predictable from the direct
interactions
Interaction modification indirect effect
– “a third species changes how a pair of
species interacts;” the third species changes
the per capita effect of one species on
another (e.g., when barnacles are present,
limpets are harder for birds to find);
difficult to predict a priori
Figure modified from Wootton (1993)
bird
limpet
barnacle
bird
barnacle
limpet
Direct & Indirect Effects
Density-mediated indirect interactions
– “indirect effects… propagated by changes in densities of
intervening species” e.g., “keystone predator effects, trophic
cascades, and exploitative competition… [as] traditionally conceived”
Approx. the same as interaction chain indirect effect
Trait-mediated indirect interactions
– “If a species reacts to the presence of a second species by altering
its phenotype [phenotypic plasticity], the trait changes in the
reacting species can alter the per capita effect of the reacting
species on other species…”
Approx. the same as interaction modification indirect effect
Werner & Peacor (2003)
Direct & Indirect Effects
Apparent competition (an example from Schmitt 1987)
Prey species:
Sessile bivalve filter feeders occur mostly in crevices
Gastropods occur on rock surfaces and graze algae
(Limited opportunities for direct competition, since neither diet nor
space requirements overlap greatly)
Common predators: Lobsters, octopi, whelks
Apparent
competition
Experiment: Continually transplanted bivalves to
maintain high densities of bivalves in sites with high
densities of gastropods
Prediction (if apparent competition operates):
Predator density will increase, gastropod density
will decrease
C
-
+
P
+
-
P
Direct & Indirect Effects
Apparent competition (an example from Schmitt 1987)
Found increased predator density and decreased gastropod density when
bivalves were added relative to control sites
Control sites
Sites with added bivalves
Figure modified from Schmitt (1987)
Direct & Indirect Effects
How important are indirect effects?
Menge (1995) reviewed 23 experimental studies of rocky intertidal habitats
that were sufficiently well replicated and long enough in duration for
indirect effects to become evident
Considered only “ecologically significant” effects (that caused at least a
10% change in the abundance of one or more species)
Found that 83 types of indirect effects accounted for 40% of the observed
changes in community structure caused by manipulations (e.g., predator or
prey removal)
Most of the indirect effects were cases of keystone predation (35%) and
apparent competition (25%)
Exploitative competition constituted only 3% of indirect effects!
Bottom-Up vs. Top-Down
Are abundances or distributions of organisms controlled by
resources (bottom-up processes) or by predation & disease
(top-down processes)?
Bottom-up view: Organisms at each
trophic level are food limited
Top-down view: Top level is food limited,
lower levels are alternately predator vs.
food limited (originated with Hairston, Smith
& Slobodkin 1960 – HSS)
Trophic cascade
C
+
+
+
H
-
+
P
Are Trophic Cascades “All Wet”?
Polis (1991), Strong (1992) & etc. argued that the idea of discrete trophic
levels, which trophic cascades are predicated on, is invalid b/c of the
prevalence of omnivory
Strong (1992) posed the question above, in part b/c omnivory appeared
more prevalent in terrestrial communities (making trophic cascades more
likely in aquatic communities)
Photo of Gary Polis from http://science.marshall.edu/fet/euscorpius/images/polis.JPG
A Terrestrial Trophic Cascade
A likely example of a terrestrial trophic
cascade (McLaren & Peterson 1994)
500 km2 Isle Royale National Park
in Lake Superior
Primary producer: Balsam fir
Herbivore: Moose (59% of winter diet
is Balsam fir)
Carnivore: Wolf (colonized island
in 1959)
Photo of Isle Royale from Wikipedia
A Terrestrial Trophic Cascade
McLaren & Peterson (1994):
“The shaded areas highlight
intervals of forage suppression
that… are closely tied to periods of
elevated moose density, which in
turn follow periods of low wolf
density (note the lags…)… these
intervals have no correspondence
to AET [climatic fluctuations]”
Figure from McLaren & Peterson (1994)
A Terrestrial Trophic Cascade
A change in behavior of a top predator cascades through a community
(Post et al. 1999)
On Isle Royale, fluctuations in North Atlantic Oscillation (NAO)
result in changes in winter snow accumulation
Annual aerial surveys show close correlation between
wolf pack size and the status of the NAO
Photo of winter wolf pack (in Yellowstone National Park) from Wikipedia
A Terrestrial Trophic Cascade
Post (1999):
“a, Increase in the mean size of
wolf packs in snowy (negative
NAO) winters…”
“b, Increase in the winter kill rate of
wolf packs with pack size…
kill rate per individual wolf also
increased during snowy winters”
“c, Decline in moose density one
year after increase in size of winter
wolf packs”
“d, Increased growth of fir trees
one year after decline in moose
density” [notice reversed x-axis]
Figure from Post (1999)
A Terrestrial Trophic Cascade
Changes in wolf behavior have ecosystem-level effects on Isle Royale
because moose dramatically influence net primary production, litter
production & edaphic nutrient dynamics
(Post et al. 1999)
Photo of moose from Wikipedia
Bottom-Up vs. Top-Down
Hunter and Price (1992) – we should always start with a bottom-up
template: “the removal of higher trophic levels leaves lower levels
present (if perhaps greatly modified), whereas the removal of primary
producers leaves no system at all”
Echoed in John McPhee’s (1998) Annals of the Former World, pg. 84:
“Break the food chain and creatures die out above the link”
Fretwell (1977) & Oksanen et al. (1981) – OFAN – proposed a
reconciliation: productivity determines the number of trophic levels that
can be supported in a community; plant productivity therefore ultimately
dictates when top-down forces could cascade back down
In general the top-down vs. bottom-up question applies to NPP, but
in principal could be asked of a variety of variables at a variety of
trophic levels.
Foundation Species
Photo from Wikipedia; definitions from Ellison et al. (2005)
“Foundation Genotypes”
Figure from Whitham et al. (2008)
Keystone Species
Keystone predator – a predator whose
activities maintain species diversity at
lower trophic levels by disallowing
competitive exclusion
(Paine 1966)
Keystone resource – first applied to
plant species that sustain frugivores
through periods of food scarcity in
tropical forests, e.g., figs
(Terborgh 1986)
Photos from Wikipedia
Pisaster eating mussel
Barbet eating fig
Ecosystem Engineers
Ecosytem engineer – an organism that creates,
modifies, or maintains habitat (or microhabitat) by
causing physical state changes in biotic or abiotic
materials that, directly or indirectly, modulate the
availability of resources to other species (Jones et al. 1994)
An ecosystem engineer has a large impact beyond simply assimilating
and dissimilating material
The definition is especially useful when applied to organisms that modify
the environment through means other than trophic activities
Photo of Clive Jones from Cary Institute of Ecosystem Studies
Ecosystem Engineers
Allogenic ecosystem engineer – organism that changes the environment
by transforming living or nonliving materials from one physical state to
another, via mechanical or other means (Jones et al. 1994)
E.g., Beaver
Photo of beaver dam on Tierra del Fuego from Wikipedia
Ecosystem Engineers
Autogenic ecosystem engineer – organism that changes the
environment via its own physical structures, i.e., living & dead tissues
(Jones et al. 1994)
E.g., Long-leaf pines
K. Harms’ photo of Pinus palustris at Camp Whispering Pines, Tangipahoa Parish, LA
Assembly Rules
Diamond (1975) coined the term for broad patterns of
bird species distributions in the Bismark Archipelago
& Solomon Islands
Wilson & Whittaker (1995; pg. 801):
“generalised restrictions on species presence or absence
that are based on the presence or absence of one or several other
species, or types of species (not simply the response of individual species
to the environment)…”
Connor & Simberloff (1979) kicked off a long and continuing debate about
assembly rules and testing for them
E. Weiher (quoted in Stokstad’s piece in Science, 2009, v. 326, pg. 34):
“I think what we’re going to find out is that assembly rules are vague,
gentle constraints”
Photo of Jared Diamond from Wikipedia
Priority Effects
Petraitis et al. (2009) provide an experimental example of priority effects
and multiple stable states in the Gulf of Maine
Ice scour can create open patches; experiments mimicked these
disturbances (rockweed stands cleared in 1996 and followed through
2005)
In sheltered bays, rockweed stands or mussel beds established,
depending on which arrived first, and were not invaded by the other
species
As required by Peterson (1984) to establish
multiple stable states, “the very same site
could come to be occupied by different,
self-replicating communities”
Figure from Petraitis et al. (2009)
Community Assembly / Coalescence
“We use the term community coalescence to refer to the development of
complex ecological communities from a regional species pool. This
coalescence depends on interactions among species availability,
physical environment, evolutionary history, and temporal
sequence of assembly.”
From: J. N. Thompson et al. 2001. Frontiers of Ecology. BioScience 51:15-24.