Transcript Competition: First look

Competition and Coexistence
First, think back to introductory ecology. “Competition
occurs between or among species iff (if and only if) some
needed resource is available only in limited supply.”
What are the resources that may be limiting?
The list includes water, mineral nutrients, light, space, and
(only a remote possibility) CO2.
Plant ecologists view competition from two perspectives:
mechanistic – how the resources are partitioned and
effect species growth and reproduction, or
based on outcome – what is the effect of interaction on
population dynamics and R0?
Considering outcomes, another carryover from ecology is the
way to categorize the variety of interactions:
Competition
--
Predation
+(or parasitism or herbivory)
Mutualism
++
Commensalism
+0
Amensalism
-0
Within competition Deborah Goldberg (U. Michigan)
subdivides ‘the interaction’ by, in effect, separating
mechanism and outcome through the intermediate (the
common factor among competing species - resource,
pollinator, herbivore, or microorganism) that ‘controls’ the
interaction.
As an example if the intermediary is nitrogen, a nitrogendependent grass population may deplete nitrogen (- effect),
but increase in response to nitrogen (+ response), with a net
negative result (-ve X +ve = -ve).
The negative net result indicates that if 2 species have
nitrogen dependence in common, the result will be resource
competition between these species.
Here’s the table that describes the forms of competitive
interaction that Goldberg identified:
What is new and different about Goldberg’s treatment is the
occurrence of ‘apparent competition’ and ‘apparent
facilitation’.
Examples of experimental studies demonstrating apparent
competition are rare. One is a study (Reader, 1992) in an
abandoned pasture near
Guelph.
3 herb species were planted
(two shown), and two
treatments (and their
Interaction) were tested:
herbivore exclusion and
neighbor removal.
With herbivore access, seedling survival was strongly affected
by competition. Within cages (no herbivore access) seedling
survival was not affected by neighbors. So, did ‘traditional’
competition occur? The apparent competition may really have
been mediated by herbivores as natural enemies.
In apparent facilitation natural enemies are once more the
explanation. A common natural enemy that is reduced by the
presence of both species in the presence of an increase of both
species would appear as cooperation between species.
Within the framework of ‘normal’ competition, one of the key
ideas is that the interaction is frequently asymmetric. Are some
occurrences of asymmetry really occurrences of apparent
competition?
Examples of apparent competition and
facilitation:
The small cactus Opuntia fragilis
grew better in the open than when
shaded with grass. This sounds like
competition for light. However, the
competition was really apparent,
because attacks by larvae of a cactus
moth were less frequent in the open.
Plant species may appear to interact with each other due to
similar habitat requirements. Annuals Minuartia uniflora and
Sedum smallii are found together in shallow depressions in
granite outcrops in North Carolina and Georgia. They
segregate according to soil depth within these depressions.
They show positive association, but are, in fact, competitors
within their specialized habitat.
Minuarta uniflora
Sedum smalii
Positive associations may also result from the occurrence of
nurse plants. Sometimes the nurse plants benefit the early
stages of the ‘nursed plant’, but its growth reverses the
apparent facilitation. The columnar cactus Neobuxbaumia
tetetzo is nursed by a shrub Mimosa when juvenile. With
further growth the cactus inhibits and replaces its nurse.
Neobuxbaumia tetetzo
Juncus gerardi
In the middle zones of salt marshes at Rhode Island the rush
Juncus gerardi (blackgrass) reduces soil salinity and increases
soil aeration, and promotes growth of other species. However,
other species have either negative or neutral effects upon the
rush.
Plants such as the sugar maple, with roots that span a large
vertical water gradient, move water from depths to the surface
where neighboring species benefit from leakage of water from
upper roots. This process is called hydraulic lift.
Two species competition experiments
There are a number of basic designs used in competition
experiments to test different aspects of interaction:
1. Partial additive designs – used to test competition effects on
a target species. Its density is held fixed, while the density
of the competitor is varied. This, and other designs, can be
described by two species abundance diagrams.
This design is commonly used to test the impact of weeds on a
crop species.
This graph presents the effect
of sicklepod (Cassia
obtusifolia) and redroot
pigweed (Amaranthus
retroflexus) on cotton yield.
Cassia obtusifolia
Amaranthus retroflexus
It is also used to assess effects of one species on another under
greenhouse conditions. For example, I used this design to test
effects of competition among fugitive species Mirabilis
hirsuta, Verbena stricta, and Solidago rigida.
My experiments were run at competitor densities of 1, 3 and 6
competitors surrounding the target plant. The general pattern
observed is of a hyperbolic response in biomass (or other
parameters of the test plant) with increasing density of
competitors.
Performance of test plant
10
8
6
4
2
0
0
2
4
6
8
Density of competitors
10
Note that this form of response is just an expansion from
looking at density effects on the growth of a single species to
an almost identical equation describing two-species
interaction.
The intraspecific density response, usually used to assess
yield in a crop, in general fits an equation of the form:
w = wm (1 + aN)-1
With two species, the effect of density of the second species
on the first requires conversion of species 2 density into
species 1 equivalents:
wi = wmi (1+aiNj )-1
Another important design is the replacement series.
Diagrammed as a joint abundance diagram…
To measure the separate influences that each species has on
the other, DeWit developed this design. There is a constant
total density of the mixture, while the abundance of the
species vary (and thus the ratio of the species).
What is typically measured in replacement series is the
relative yields of the species in mixtures compared to those in
monoculture at the same density…
Relative yield of species i
= (yield per unit area of i in mix)
yield per unit area of i in monoculture
Relative yield of species j
= (yield per unit area of j in mix)
yield per unit area of j in monoculture
If the species differ in their density responses, the combined
measure is used…
Relative yield total (RYT) =
 of relative yields for species i and j
As long as both species in the comparison are growing at
densities within the range that generates a constant final
yield, an RYT > 1 indicates two things about the species:
1. At least at the density tested, one of the two species has an
advantage; it is a superior competitor.
2. At least at the density tested, the RYT can be taken as
evidence of some niche separation. And
3. It is evidence of greater yield than an equivalent area
planted in two monocultures; this is of agricultural interest.
Other indices have been proposed to avoid some of the
problems with RYT, but require knowledge of the full density
responses of both crops used in the mixture.
The third design is the additive series.
And the last, very similar, design is the complete additive
design
In these two designs the densities Ni and Nj can be varied
independently. This makes it possible to determine/fit the
isoclines of the theoretical graphical models with real data.
The usual method is to use input-output measures, i.e. seed
planted and seed production by plants.
Recall the equation for density-dependence in a single
species:
Nt+1 = smNt(1+aNt)-1
where sm is the maximum seed production of individuals
planted at very low density. To generalize this equation to
competition among 2 or more species, the equation for one of
the species, i, has a sum (taken over j = 1 to n) of self and
other species effects…
Ni,t+1 = smiNi,t(1+ αijNj,t)-1
To normalize the interaction coefficients to ‘equivalence’ with
the competition coefficients as slopes of isoclines in basic
plots, each must be divided by ii.
In an experiment using this approach for the competition
among 3 species – Lolium rigidum, Vulpia bromoides and
Trifolium subterraneum – phase plane diagrams and directions
of population growth indicated:
1. Competition between the clover (Trifolium) and either of
the grasses leads to a stable equilibrium, but
2. An apparently stable interaction between clover and Lolium
becomes unstable when invaded by Vulpia. Under typical
Australian drought conditions, it appears that Vulpia will
eventually displace Lolium.
The species:
Lolium rigidum
Trifolium subterraneum
Vulpia bromoides
The phase plane diagrams:
Zero isoclines for
Trifolium with either grass
are almost perpendicular.
That means they only
marginally influence each
other’s densities.
This probably indicates
niche separation in the use
of nitrogen as a resource.
Trifolium is a legume
(fixes nitrogen), while the
grasses require it.
However, the phase
plane diagram of the two
grasses in competition
indicates eventual
replacement and loss of
one.
Multi-species competition
Equations attempting to incorporate all species in a system
into models become prohibitively complex.
The standard method for exploring complex systems is use of
removal experiments. One species is removed, and its effects
on other species studied by comparison with results in its
presence.
The results of removal experiments is not always intuitive.
There are ‘hidden’ interactions that affect observed results.
Norma Fowler’s experiments in North Carolina looked at two
particular species in the context of a grassland community:
Rumex acetosella and Plantago lanceolata.
The effect of removal of plantain (Plantago) was conditional
on the presence/absence of the sheep’s sorrel (Rumex). When
sorrel was absent, a whole suite of winter annuals increased in
abundance. However, …
When sorrel was present, it increased in abundance, and the
winter annuals did not.
There are also 3rd party effects: when one species is removed,
another may decrease, rather than increasing. It’s the result of
both interacting with a 3rd species.
Again the example considers two species and the community
of other grasses present, here in Washington subalpine
meadows. The monocots directly assessed were Carex
spectabilis, which was the species removed, and Festuca
idahoensis.
Removal of the Carex in plots where it was a dominant
species resulted in increased abundance of the Festuca, but
decreased abundance in 4 other grasses.
The explanation: Festuca was the important competitor for the
other grasses, and once the effect of Carex on it was removed,
it depressed the other grasses.
Spatial factors are frequently important in removal
experiments…In the short term, responses may be strongest in
those individuals close to the gaps created by removal.
When studying multiple species competition relationships,
when species’ positions in the hierarchy of competitive
dominance are consistent (i.e. A > B > C…) the relationship
among species is described as transitive.
That is not always the case. There are many instances where A
> B and B > C, but C > A, for example.
The other aspect that is important in determining the result of
competition is spatial. Effects on an individual plant occur
when competitors are located within its neighborhood area
(or area of influence). Plants beyond that zone can be
disregarded.
The spatial dimension includes height in addition to lateral
distance.
When Douglas fir (Pseudtsuga menziessi) seedlings compete
with shrubs, regression models indicated little influence of
shrubs with a relative height less than 1.25 (relative to the
target (fir) height being taken as 1).
In the same set of experiments, the distance between the fir
seedling and its competitor(s) is also important…
The effects of competitors increase as the distance from the
seedling stem decreases.
Effects of other organisms on plant compeititon
Among the types of organisms that can affect competition are:
herbivores, pathogens, and symbiotic microorganisms
(particularly mycorrhizae).
It should be apparent that herbivory can alter the competitive
balance. While studying competition on prairie grasslands, I
ran a ‘cafeteria’ experiment on the preferences of prairie voles
for C3 versus C4 grasses.
Voles preferentially consume C3 species, probably because of
the anatomical structure and distribution of nutrients in C4
species.
How important that preference is in determining the
dominance of C4 grasses on northwest Iowa prairies is difficult
to determine.
Herbivore Effects
The beetle Gastrophysa viridula significantly reduced leaf
area and weight of Rumex obtusifolius when the plant was
growing with grass but not when it was growing without
competition. Herbivore activity and effect may be conditional.
Pathogen Effects
Pathogens may exert their effects directly, e.g. by affecting the
fitness of infected plants, but may also have influence by
shifting the burden of herbivory.
Many endophytic fungi produce toxins (alkaloids, cyanogenic
glycosides) that affect herbivore preferences, and therefore
shift the competitive balance.
Endomycorrizal fungi can act as mutualists, aiding
particularly in nitrogen and phosphorus acquisition, but can
also move nutrients from one plant to another.
There is a huge variety of mycorrhizal species. It seems,
experimentally, that plants are best adapted to their local
species…
Clover biomass is plotted against different species of grass
from a Canadian pasture (each with their own fungal flora).
Clovers were different genotypes from areas with each of the
grasses. In each case the best growth occurred with the ‘home’
fungal flora.
There are a number of newer models for competition that take
a different view for competition. Going back to the basic
Lotka-Volterra model, the older views are equilibrium
models; the newer views are various forms of nonequilibrium models.
The first of these models, developed by Alexander Watt, was
a wave regeneration model. Watt studied cyclic regeneration
in Calluna heath in uplands of Scotland. The cycle:
Calluna (heather) is initially competitively dominant,
excluding lichens. As it ages, it can be invaded by lichens
(Cladonia). Its mats are broken by winds, leaving patches of
bare soil. Arctostaphylos (bearberry) invades the bare patches.
Bearberry patches are invaded from the edges by Calluna,
more-or-less slowly replacing the bearberry, completing the
cycle.
Diagrammatically:
A second type of model, collectively called lottery models
with a name coming from Peter Sale, suggests that it is chance
(stochastic) events that determines the composition of the
community. Usually this ‘chance’ is the result of the
recruitment process.
The diagram in the text is slightly different than the situation
on the prairie…
When time, as well as spatial factors, influence who ‘wins’ so
that species coexistence is based on varying environmental
conditions, the effect of variation is described as a storage
effect. The species coexist because each has the advantage
under one set of conditions.
There is a completely different class of models, first proposed
by Steve Hubbell, that have their origin in the genetic models
of Moto Kimura. They are called neutral models. In these
models species and individuals are considered equivalent. The
structure and composition of the community, in the end, is the
result of random events, long-term speciation processes, and
the balance between local colonizations and extinctions.
Hubbell developed the model from his experience in detailed
study of the structure of tropical forest communities in Costa
Rica.
Hubbell’s neutral model fits his data from the tropical
rainforest quite well (as do other relative abundance models).
However, the basic assumption that species are essentially
exactly equivalent has been considered a severe problem.
Experimental evidence, as well as the competitive exclusion
principle, suggest that equivalent species don’t coexist.
But Hubbell’s model is also based on random events, and is a
non-equilibrium model. Is the competitive exclusion principle
limited to the equilibrium situation?