Nonequilibrium models of the maintenance of diversity

1
Non-equilibrium models of the maintenance of
diversity
Hubbell and Foster (1986) Theoreticians abhor
unique events! - Theoretical community ecology
has not paid sufficient attention to historical
explanations of ecological phenomena (fear loss
of predictability?) - Examined role of chance
colonization and history in structuring tropical
tree communities. - Fundamental question How
do you account for the maintenance of many
similar species in species-rich communities (e.g.
corals on reefs, plankton in lakes, trees in
forests)
2
The problem with classical equilibrium theory
Can characterize these theories as niche
assembly models of community structure Classic
models based on competition and predation, mostly
have predicted the coexistence of only small
numbers of species. How do these mechanisms
scale up to explain the persistence of numerous
species in communities? e.g. Hutchinsons
Paradox of plankton and Homage to Santa
Rosalia (1959) Question still largely
unresolved and remains the Grail of community
ecology Most attention given to communities of
producer organisms (plants and coral reefs)
3
Take the case of plants (for a change)
- Biotic environment of a rooted organism is
inherently statistical in nature, and for a
species rich-community would also consist of a
large number of neighbor species whose identity
is also unpredictable - Complexity and
unpredictability of the biotic neighbourhood
limits the constancy in direction of selection
that leads to specialization. - Similarity in
traits seen among many tree species might reflect
convergent generalization in response to the
long-term average diffuse selective forces
imposed by the sum of all competitors
4
Hubbell random drift model If species are
competitively equivalent generalists Then
Diversity a consequence of regional tree species
richness and availability of potential immigrants
( climate, historical biogeography, local
dispersal, and speciation on regional scales)
Hubbells drift model is not the only
non-equilibrium model A similar idea to Hubbell
is the lottery model developed to explain
maintenance of diversity in systems such as coral
reef fish (Chesson and Warner 1981). According
to this model, openings such as territories
available for fish, are filled by colonists at
random.
5
Related to the lottery model is the storage
effect of Warner and Chesson (1985). - More
deterministic than a lottery, but depends of how
conditions that favor different species
fluctuate - Imagine a desert scrub system. All
species have some conditions under which their
populations can grow good years. Different
species benefit in different years -Effects of
good years can be stored by survival of long
lived adults through unfavorable times or by
survival of propagules (e.g seeds in the seed
bank awaiting favorable conditions for
germination).
6
Pake and Venable (1996) looked at dormancy and
seed germination of Sonoran Desert annual plants.
- Many species produced a super-annual seed
bank - Germination responses to temperature
variation were species-specific - Species at low
densities therefore may occasionally experience a
favorable year in the absence of competition
(because germination responses are not completely
correlated among species) - This allows a
low-density species to sometimes attain high
population growth rates and re-invade a
community...
7
Intermediate Disturbance Hypothesis (Connell 1978)
Frequency and intensity of various kinds of
abiotic disturbances affect patterns of
diversity. Disturbance is critically important
in structuring communities because it can prevent
competitively dominant species from excluding
others. Weak/infrequent disturbances
insufficient to prevent competitive exclusion
Intense/frequent disturbances exclude species
sensitive to disturbance Highest diversity might
therefore be expected at intermediate frequency
or intensities of disturbance
8
Intermediate Disturbance Hypothesis
9
Wayne Sousa (1979). Most famous Test of Connells
Intermediate Disturbance Hypothesis
How does disturbance to boulders affect diversity
of algae? Disturbance is (reasonably)
predictable and frequent, and post-disturbance
colonization rapid enough to make experiments
possible - Waves in winter storms turn over
boulders exposing new substrate for colonization
and smothering and killing algae on the underside
of the rock. What influences local disturbance
rate??
10
Frequency of disturbance determined by rock size
(and therefore force needed to move
it). Classify rocks according to force
categories.
Found that Recently turned over boulders are
colonized by a green alga (Ulva) and
barnacles At an intermediate time
post-disturbance boulder space is shared between
Ulva and several mid-successional red
algae. Undisturbed boulders become dominated by
the late successional alga Gigartina
11
Mid-successional algae most abundant at
intermediate disturbances
Note Numbers above columns are sample sizes x
axis pull is force required to move a boulder
Figure shows density of mid-successional red algae
12
Sousa also performed Connell-style experiment
If disturbance is confounded with boulder size
then size may be correlated with other factors
that influence recruitment and survival of algal
species (rem Island biogeography) Sterilized
boulders, cemented half in place and looked at
colonization Unstabilized rocks dominated by
Ulva after 18 mo (early successional
species). Stabilized rocks had diverse cover of
red algae (mid-successional species).
13
Application to IDH to forest communities Disturba
nce in forests is from treefalls (blow-downs and
landslides) If IDH operates what tree
communities would be expected to have highest
diversity? What determines the tree-fall rate in
forests?? Problem with IDH for tropical
forests Few species are specialized on
disturbances (10-20 ). Most species present
before the disturbance occurs - recruitment
limitation more important than disturbance rate?
14
Recruitment limitation
Failure to arrive, or to establish at a site
suitable for onward growth. This encompasses
both failure of dispersal (dispersal limitation)
and absence of suitable microsites for seedling
establishment (establishment limitation). Establi
shment limitation may be imposed by either
absence of suitable abiotic conditions (e.g
exposed soil, sufficient light) or biotic
conditions (e.g. avoidance of density dependent
processes) Good evidence for both dispersal
limitation and establishment limitation in many
community types
15
Example of recruitment limitation in a marine
system (Birkeland 1982)
Crown of Thorns starfish Acanthaster planci
16
What determines when crown of thorns is abundant?
Adult Acanthaster outbreaks occur, suddenly and
irregularly for at least 2 centuries at sites
remote from human populations. Coral cores
suggest that is a long-term dynamic Outbreaks
have become more frequent and long-lasting in
recent years Understanding what causes these
outbreaks is important because adult populations
reach very large numbers and munch prodigious
quantities of corale.g. 500,000 removed in 1
year from Tutuila, American Samoa Individual
COTS can cover 20 km2 of coral a year.
17
Hypothesis Outbreaks of COTS represents periodic
escape from recruitment limitation Birkeland
figured out that outbreaks occur 3 years after
typhoons bringing heavy rain (but do not follow
dry typhoons) Explanation appears to be that
terrestrial run-off from extreme storm events
provides sufficient nutrients to stimulate
phytoplankton blooms of sufficient magnitude to
produce enough food for Acanthaster
larvae. Increased survival of larvae results in
high settlement rates on reefs across a large
area and high adult densities 3 yr later. Adult
COTS produce 65 million eggs! Small increases in
larval survival probability can result in rapid
increases in recruitment
18
Recruitment limitation in context of other traits
Remember example of Drosophila competition -
colonization trade-off??
Many examples for plant communities too
(succession lecture and discussion) Tilman
(1994) provides a spatial model that shows that
if a strict trade-off exists between dispersal
and competition then an infinite number of
species can coexist competing for a single
limiting resource. What would happen without
this trade-off?
19
Species differ in R value for N - the limiting
resource in young successional grasslands
Support for competition-colonization trade-off
for plants at Cedar Creek, MI
20
What about colonization alone (dispersal
limitation)?
Hurtt and Pacala (1995) simulation models to look
at how dispersal would slow competitive
exclusion. Rank species according to competitive
ability but give all species equivalent and
restricted dispersal ability. Limits to
dispersal can dramatically slow population and
community dynamics in species rich
communities Important in diverse communities
because increased species richness causes each
species to become proportionally rarer, which in
turn causes each species to become more dispersal
limited. As a consequence most recruitment sites
are won by forfeit rather than by the best
possible disperser
21
So, how dispersal limited are communities?
Long-term seed trapping studies can be used to
see how frequently seeds arrive at potential
recruitment sites
22
Hubbell et al. 1999
Study of seed capture to 200 seed traps over a 50
ha area 216 plant species captured in the traps
over 10 years Of these, only 7 dispersed gt1 seed
into gt75 of traps 50 of species dispersed
seeds to gt7 traps Dispersal limitation is very
strong in species rich communities. Dispersal
needs to be taken into account before considering
any other species interaction Is dispersal alone
sufficient to explain diversity patterns?