Niche and neutrality debate in community ecology

1
Niche and neutrality debate in community ecology
By Sandra P. Galeano
2
Diversity maintenances in communities

• Explaining maintenance of species diversity
  within
• communities one of the most fundamental
• questions in ecology
• Patterns from mechanisms? mechanisms by
• patterns?
• Niche - neutrality debate

3
Niche-neutrality debate some history

• The classical niche paradigm
• Grinnell (1917) Geophysical spatial unit
  with the niche relationships of the California
  thrashers
• Gause (1934) Competitive exclusion
  principle
• From here, discussion relies almost exclusively
  on competition
• Hutchinson (1957) Niche hypervolume,
• Ecological niche sufficient and necessary to
  explain coexistence!

4
The competitive niche paradigm

• D. Tilman
• Resource based theory R (1982,1988)
• Life history trade-offs, essential for
  coexistence
• Arbitrary number of species could coexist based
  on
• fugitive species concept of Hutchinson!
• Hurtt Pacala (1995) Recruitment and dispersal
• limitation competitive exclusion infinitely
  delayed
• S. Hubbell (1990s)
• Not all species clearly niche differentiated.
  Competitive exclusion or character displacement
  not always evident Ecological equivalent
  hypothesis and neutral theory

Is ecological niche sufficient to explain
coexistence?
5
Neutral theorys in Biology

• Neutral theory in Biology did not start with
• Hubbells theory

2001
1960s
6
Niche-neutrality debate some history

• The unified neutral theory of biodiversity and
  biogeography (UNT)
• S. Hubbell (2001)

Annual survival
after observing
Trade-off (gtsurvival shade sps-low growth in
sun) Species undistinguishable
Growth in gaps
Hubbell Foster 1992
7
Hubbells neutral theory of biodiversity and
biogeography

• Distribution of species in an environment
  stochastic processes operating at
• regional scale
• local scale

Births
8
Hubbells neutral theory assumptions

• The zero-sum ecological drift
• a) Community fixed number of individuals -
  saturated
• b) each death replaced by a birth at random by
  any individual
• Ecological equivalence Individuals of all
  species are demographically similar (equivalent
  probabilities of birth,
• death, dispersal, speciation)very
  criticized!!!
• Null random and it is used as statistical test
  of empirical data
• Neutral Assumes functionally equivalence - uses
  parameters

9
Hubbells neutral theory

• What patterns result from the assumptions that
  all species exhibit same demographic
  stochasticity and only ecological drift,
  dispersal, and speciation occur?
• How similar are patterns predicted from UNT and
  nature?

Abundance


Species sequence
matches almost all distributions of relative
abundance
Hubbell 1997
Magurran 1988
10
Hubbells neutral theory
Hubbell 2006

• Prediction of UNT
• fits really well species
• abundance distribution of a
• tree community at Borneo

Hubbell 2001

• Prediction of UNT
• fits really well relative species abundance at
  BCI, Panama

DESPITE RADICAL ASSUMPTIONS!
11
Niche and neutrality debate

• Niche vs neutrality - the extreme perspective
• Is this way of thinking appropriate
  productive?
• What empirical data suggest?

vs
Each finch has its niche or niches are not
important at all
12
Niche and neutrality debate in community ecology
study cases

• Species abundance distributions
• Assessing the fit of a neutral model to relative
• abundance patterns
• Identifying the mechanisms for relative abundance
  
• Spatial distribution (environment or dispersal?)
• Local scale
• Other studies

13
Species abundance distributionAssessing the fit
of a neutral model

• A test of the unified neutral theory of
  biodiversity
• McGuill (2003a)
• Used data from BCI tree data set (same site than
  Hubbell)
• The log-normal and zero-sum model were fitted to
  the data
• Used maximum-likelihood
• methods and measured eight
• goodness of fit

14
Species abundance distributionAssessing the fit
of a neutral model

• The lognormal beats the zero-sum on all goodness
  of fit!
• - However, success of lognormal does not mean
  that community structure is random. Many other
  models might explain it
• - Low number of parameters and greater parsimony

McGuill 2003a
15
Species abundance distributionAssessing the fit
of a neutral model

• Why did the results differed from Hubbell 2001?
• Fitting the data for a year or a cutoff in
  minimum
• tree size different than Hubells data?
• The study found some issues when fitting the
  ZSM
• ZSM predicts it will take thousands of years to
  reach local
• equilibrium (longer than time for environment
  to be constant)
• ZSM methods produce too many rare
• species so distribution must be stopped
  
• better prediction than log-normal

Sample size!
McGuill 2003a
16
Assessing the fit of a neutral model

• Curve fitting is a weak test- McGuill (2003b)
• - More than one mechanism produce any given
  distribution
• There is not such a thing as a curve that fits
  best
• (depends on goodness of fit measure used)
• McGuill suggests better methods to test theories
  (e.g.
• constancy in parameters, correlations, temporal
  and
• spatial scale of exchangeability, etc)
• So fitting the curve seem to be
  counterproductive as it
• does not reveal the mechanisms!

17
Niche and neutrality debate in community ecology
study cases

• Species abundance distributions
• Assessing the fit of a neutral model to relative
• abundance patterns
• Identifying the mechanisms for relative abundance
  
• Spatial distribution (environment or dispersal?)
• Local scale
• Other studies

18
Species abundance distributionIdentifying the
mechanisms

• Non-neutral patterns of sps abundance in
  grasslands
• Harpole Tilman (2006)
• Three communities
• Minnesota (CDR oldfield) 60 years abandonment
• Minessota (CDR experiments) 5 years
  experimentally assembled community (no regional
  dispersal)
• Kansas (KNZ) relict fragments
• Species traits to predict identity of
• dominant vs rare species (McGill 2003)
• N (nitrogen)

19
Species abundance distributionAssessing the
mechanisms

• Distribution of species abundance consistent with
  neutral
• theory?

• - Data from CDR oldfields
• (percent cover data)
• Used an implementation of Volkovs solution to
  fit ZSM (Volkov et al. 2003)
• -Species abundance distribution was fit well by
  both the ZSM and lognormal

Harpole Tilman 2006
20
Species abundance distributionAssessing the
mechanisms

• Are species equivalent?
• - Three communities
• - R index inversely related
• to competitive ability for N
• - Correlations and likelihood
• of mean correlation by chance
• - Lower R species had greater
• abundance than higher R

Relative cover
Relative biomass
Harpole Tilman 2006
R
21
Species abundance distributionAssessing the
mechanisms

• Are species equivalent?
• - Three communities
• - R index inversely related
• to competitive ability for N
• - Correlations and likelihood
• of mean correlation by chance
• - Lower R species had greater
• abundance than higher R
• - Mean correlation differs from
• 95 CI neutrality

Harpole Tilman 2006
22
Species abundance
distribution
Assessing the mechanisms
Species abundance change in a neutral manner with
environmental change?

• -Data from previous N addition
• experiment (CDR experiment)
• - Biomass
• - Correlations and likelihood
• of a relationship arising
• under neutral assumptions
• - Correlation was negative at low
• N lower R sps more abundant.
• At high N, correlation becomes
• Positive so higher R abundant

Harpole Tilman 2006
23
Species abundance distributionIdentifying the
mechanisms

• Tests supported niche mechanisms as primary
  drivers of grassland communities
• Even for CDR oldfields (species abundance was
  consistent with neutral theory)
• Supports McGuill (2003), curve fitting might lead
  to incorrect assumptions
• But, greatest variation in R
• abundance correlation inCDR oldfields
• Dispersal still influence community,
• so conclusion?

Harpole Tilman 2006
24
Niche and neutrality debate in community ecology
study cases

• Species abundance distributions
• Assessing the fit of a neutral model to relative
• abundance patterns
• Identifying the mechanisms for relative abundance
  
• Spatial distribution (environment or dispersal?)
• Local scale
• Other studies

25
Spatial distribution environmental or dispersal?

• Neutrality, niches, and dispersal in a temperate
  forest
• understory Gilbert Lechowicz (2004)
• Gault Nature reserve, Montreal
• - Design decoupling distance and environment!
  non-correlated for sites selection (GIS and
  Mantel tests)

26
Spatial distribution environmental or dispersal?

• Neutral theory predicts a decrease in community
  similarity with distance (effects of dispersal
  limitation)
• Observed data does not follow neutral prediction
  (environmental-distance effects decoupled)

Gilbert Lechowicz 2004
27
Spatial distribution
environmental or dispersal?

• Measured soil moisture, soil nutrients, and other
  22 physiographic variables
• CCA analyses to identify species-specific niches
  and dispersal patterns
• Environmental explains plant distribution better
  than dispersal (neutral), but dispersal partially
  explains dispersal, so it is also important

Gilbert Lechowicz 2004
28
Spatial distribution
environmental or dispersal?

• From CCA analyses determined relative importance
  of variables explaining variations in species
  distribution
• High score on one axis species within group
  differentiate ecologically along that axis
  (distinct niches).
• Shows evidence for niche structuring

Gilbert Lechowicz 2004
29
Spatial distribution environmental or dispersal?

• Tests supported environment (niche mechanisms)
  as
• primary drivers for plant distribution, but
  dispersal
• (neutral mechanism) was also important
• Apparent neutral patterns of species
  aggregation
• previously detected in the reserve
• due to distance-environment correl.
• Niche differenciation occurred at
• both functional or phylogenetic level

Gilbert Lechowicz 2004
30
Niche and neutrality debate

• Niche vs neutrality the extreme perspective
• Niche and neutral processes are not mutually
  exclusive, therefore it is more important to ask
  whether niche mechanisms are significantly more
  important than an expectation from neutrality

vs
Each finch has its niche or niches are not
important at all
Gilbert Lechowicz 2004
31
Niche and neutrality debate in community ecology
study cases

• Other studies
• Diversity patterns
• Beta diversity, species-area relationship,
  local vs regional
• Evolutionary characteristics
• Speciation rates, phylogenetic patterns
• It is a growing field, emerging as one of the
  most often tested with field data and evaluated
  with models (Clark 2008)
• Sesile organism biased
• but more publications with
• animals from 2009!!