Transcript Species diversity: patterns and causal hypotheses

Biodiversity
Ricklefs’ definition:
Biodiversity is “a measure of the variety of organisms within
a local area or region, often including genetic variation,
taxonomic uniqueness, and endemism.
Ricklefs is slightly more precise when he defines diversity as
an ecological measure…
Diversity is:
a) the number of taxa in a local area or region
or
b) A measure of the variety of taxa in a community that takes
into account the relative abundance of each one.
Definition a is usually called species richness. As a count of
the number of species, it does not take into account relative
abundance. It would claim the diversity of a community with
100 of species A and 1 each of species B,C,D, and E was 5,
the same as one with 20 individuals of each species.
Depending on the details of the measure used, definition b
may be called a dominance index or, if information theory is
used to generate the index, it is called Shannon’s or the
information theory diversity index.
The dominance index for a community is calculated as:
D
1
s

i 1
pi 2
The Shannon information theory diversity index is calculated
s
as:
H    pi log 2 pi
i 1
Since H is an index related to logs, two ‘corrections’ have
been suggested for it:
a) Instead of using H, compare it as a ratio to the maximum
possible diversity for the community (if all species were
equally abundant, called H’). This ratio is called evenness.
b) Make the value more proportional to the number of species
by using the H as the exponent of an expression that doesn’t
have an accepted name. The reported diversity is then 2H.
Note that your text (pg. 417) suggests using logs base e. The
result is comparatively identical, but isn’t an information
theory measure.
Here’s a sample calculation of the two diversity indices for a
community containing 5 species:
Species richness = 5
Relative abundances:
Species
# individuals
A
15
B
20
C
75
D
4
E
__8___
Total
122
pi
.123
.164
.615
.033
.066
D = 1/0.425 = 2.352
H = 1.142 (using ln rather than log base 2)
pi2
.015
.027
.378
.001
.004
0.425
Obviously, a more diverse community has more species and
no single (or a few) species predominant in the community.
But what determines how many species we are likely to find,
and what determines how many species there are in the
world?
The answer to the second question is fairly straightforward…
There are about 2 million species that have been officially
named. Some groups are well known, others have only a
small fraction named. You can probably guess which are well
known:
Birds – there are about 9000 species. Half had been named
by the middle of the 19th century. Today a few species are
named each year.
Mammals – there are about 4000 species, and less than 10
are added each year (almost all small rodents).
Are they the most diverse animal fauna? NO!
Globally, the most accepted estimates of total taxonomic
diversity expect that if we could find and name them all,
we’d have somewhere between 10 and 30 million species.
The majority of the total are insects, and among insects the
largest number are beetles. J.B.S. Haldane wrote that “The
creator has an inordinate fondness for beetles.”
About 20% of all species are beetles, and we find new
species frequently and in large numbers…
Terry Erwin has collected beetles from trees in tropical
forests for many years. In one series of studies he used
insecticidal fogging to collect insects in the Tambopata
Reserve in Panama. The results are staggering.
The species overlap between two plots only 50m apart in his
study area was only 8.7% (i.e. 91.3% of the beetles were
different over this short distance).
There were 163 species of beetles unique to one species of
tree, Luehea seemannii, in its canopy. Since beetles make up
about 40% of all insects, Erwin estimated that this tree
canopy harbored ~400 unique species of insects (160 beetles
and 240 other insects). What about the other parts of the tree
– trunk and roots? We can only guess how many others may
live there. Estimates (E.O. Wilson 1992) suggest about twice
as many species in the canopy as on and in the ground. So,
there are maybe 600 species of insect unique to this tree
species.
There are an estimated 50,000 tropical tree species, so…
There may be 30,000,000 tropical insect species.
What about the rest of the world? Tropical rainforest is the
most diverse biome on earth. Many fewer species would be
added by expanding the range of habitats included.
There are also errors of unknown size in each step of the
estimation process.
So, with no better information to make an estimate of total
global diversity more precise or more accurate, the loose
estimate of from 10 – 30 million is a useful start.
While it is generally accepted that tropical rainforest is the
most diverse biome on earth, there are also indications of a
large scale pattern in diversity globally.
In general, the diversity of species
declines with latitude, i.e. as you
move from the tropics toward the
poles. Your text uses the number
of ant species to show the pattern…
Amazon rain forest
Many taxa have been surveyed in North America. Mammals
also fit the latitudinal pattern, so do lizards. In each case
equal areas are compared, so it isn’t just that the range is
larger at lower latitude…
Mammals –
lizards
Ecologists have generated a number of different (but in many
cases not completely independent) hypotheses to explain
the latitudinal gradient in diversity. We’ll consider the
hypothesis in turn. They are:
1. Evolutionary time
8. Competition
2. Ecological time
9. Disturbance
3. Climatic stability
10. Predation
4. Climatic predictability
5. Spatial heterogeneity
6. Productivity
7. Stability of primary production
These mechanisms are likely working at different spatial
scales.
At the geographical or regional scales:
evolutionary time
ecological time
climatic stability or predictability
hypotheses invoking increased resource partitioning
spatial heterogeneity
competition
productivity and stability of productivity
At the local scale:
disturbance
predation
Evolutionary time – based on the assumption that diversity
increases with the age of a community.
In the northern hemisphere, relatively recent glaciations
mean that temperate and polar communities are very recent
compared to tropical communities. This is argued to explain
their relative paucity in species.
However, we now know that, though tropical rainforest
persisted through the 4 cycles of Pleistocene glaciation, the
community was not untouched. Rainfall patterns changed
during glacial epochs, and continuous tropical rainforest was
fragmented during glaciation (in the Amazon Basin into at
least 7 separate continuing forest areas; the areas between
were probably savannah.). So, there are old and relatively
recent communities in tropical rainforest.
Here’s the simple view of what the evolutionary time
hypothesis suggests:
Ecological time – this hypothesis suggests that the time since
glaciation has been too short for dispersal of species to
occupy their full range of habitats (and latitudes).
When glaciers receded, there was newly open habitat. How
rapidly can species migrate/disperse to occupy suitable
places? Dispersal capacity for most species suggest that this
is at most a minor factor.
Climatic stability – Here reference is made to the amplitude
of the seasonal cycle of climate.
If there is little seasonality, species can live there with little
need for broad tolerance capability (a broad niche with
respect to physical climatic variables). However, if there is
high amplitude seasonal variation, either broad tolerances or
other adaptations are necessary. If niches are broad, how
many species can be crowded in? Fewer than if niches are
narrower.
Other adaptations? These can include hibernation, diapause
(loosely, insect hibernation), dormancy (plants), and,
possibly, the evolution of migration.
Along the same gradient (physical conditions or some
resource) you can pack more species with narrow niches…
3 spp.
Species
Abundance
breadth
Resource or climatic gradient
6 spp.
Species
Abundance
Resource or climatic gradient
Climatic predictability – If a climate has high amplitude
cycles, but those cycles occur predictably each year, then
adaptations can be selected to manage the problem.
Most (if not all) of the “other” adaptations to climatic
stability may well fit better here, as adaptations to predictable
cycles.
Species may time important phases of the annual life cycle to
fit environmental conditions. For example, in the desert of the
American southwest, there are two periods when rainfall is
more likely…
Here’s a comparison between rainfall at Omaha, Nebraska
and Phoenix, Arizona.
Note the difference in
the scales along the yaxes.
In the southwest desert there are two groups of plants. One’s
seeds germinate in the cool-wet (winter annuals), and the
other’s germinate in warm-wet, and flower after summer
flash floods.
Spatial heterogeneity – there is a strong correlation between
the structural complexity of a habitat and the number of
species it holds. Text figure 23.5 shows this for birds from
deciduous forest in eastern North America…
Just to show you this wasn’t a carefully chosen habitat, the
same relationship is found for birds over a wide range of
habitats in both North America and Australia…
If there are more possible specializations in habitat
conditions (indicated by such measures as foliage height
diversity), then more species may be able to co-occur by
specialization on a particular part of the ‘gradient’.
sp. 1
sp. 2
sp. 3
Fewer resources or
possible habitat
specializations, so fewer
species
Species
Abundance
Resource
More resources or…, more species
sp. 1
sp. 2
sp. 3
sp. 4
Species
Abundance
Resource
sp. 5
sp. 6
Productivity – More productivity means more food
available, and the potential for greater specialization. Thus,
everything shown on those last graphs applies here, as well.
One way of seeing this is to look at how a graph of resources
available can be subdivided into equal areas, and how many
such areas pack in when there are more resources available.
Your text (fig.23.3) shows that productivity and habitat
complexity interact to some degree. High productivity
without a means to specialize (complexity) will simply lead
to larger population sizes rather than increased diversity.
Sometimes productivity is estimated by using potential
evapotranspiration. This is a measure that combines
evaporation from the soil surface and transpiration. It
indicates energy input into a system. There is a pattern
relating PET to species richness in a number of major
taxonomic groups. Ricklefs shows the relationships for North
America.
Stability of primary production – Species can only
successfully partition resources finely if the resources do not
vary too substantially in abundance over time. Some call this
hypothesis one of temporal heterogeneity to parallel the
spatial heterogeneity hypothesis, but clearly with an opposite
result on diversity.
In the attempt to avoid predation on fruits and/or seeds, some
plants attempt to become ‘unpredictable’ in providing
resources by adopting a strategy of masting, producing a
heavy crop one year, but very little the next. That
unpredictability should reduce the diversity (and number) of
granivores or frugivores feeding from the plant.
Competition – this hypothesis has a lot of overlap with
others that invoke small niche breadth as being key to high
diversity. For example, it is believed that tropical forest
species are highly K-selected. Competition is fierce, the
populations are near their equilibria in the face of both intraand interspecific competition. There is strong selective
pressure to reduce niche breath to a region of niche space
where a species is superior. The resulting narrow niches
allow more species to be packed into the same resource
space.
Just the opposite – broader niches, lower diversity – would
be expected where r-selection predominates, i.e. harsh, high
latitude habitats.
Disturbance (and the intermediate disturbance hypothesis)
– Disturbances reduce the density of individuals, and thus the
intensity of competition. You can consider this to be the
inverse of the competition hypothesis in many senses.
The intermediate disturbance hypothesis says that when
disturbances are rare, the community becomes filled with the
K-strategists. When disturbances are frequent, those capable
of rapid growth to recover following the disturbance should
predominate, i.e. r-strategists. At an intermediate frequency
and intensity of disturbance, members of both groups may be
present, and thus diversity may be maximized. Here’s a figure
abstracting this idea…
When disturbances occur frequently, populations can never
grow to an abundance where resources become limiting.
Thus, competition among the species cannot be important.
However, species that are slow to mature and reproduce
cannot tolerate frequent disturbance, i.e. K-strategists.
The species that can tolerate frequent disturbance are those
that mature and reproduce rapidly. Those species, rstrategists, also tend to have high rates of dispersal, and are
therefore likely to ‘find’ disturbances soon after they occur.
If disturbances occur rarely, then populations do reach a size
where resource limitation is important. Thus, competition
among the species occurs, and dominant competitors can
exclude poorer competitors from the community.
Species diversity is lower due to the loss of these inferior
competitors.
But, at an intermediate frequency of disturbance (and each
place I’ve talked about frequency you could substitute
intensity) both r- and K-types may persist together, and
diversity should be higher.
There is evidence both supporting and not supporting the
intermediate disturbance hypothesis…
First, evidence that doesn’t work: the effect of burning on
plant species diversity on Konza Prairie in Kansas. Here,
species diversity declines with increasing frequency of
burning, but there is no intermediate peak…
But here is evidence that supports the hypothesis. It was
initially constructed by Joe Connell for the intertidal. Here it
was applied by Wayne Sousa to colonization of shoreline
boulders by green algae. Each line is a species. Ulva, the line
that starts first, is a pioneer. Note that it has disappeared by
the end. Both the species that arrive late and Ulva are present
at the middle of the sequence…
Sousa also knew that wave action (the main disturbance
force), would be most likely to roll small boulders (frequent
disturbance), least likely to roll large ones, and roll medium
sizes and intermediate number of times. What was the
diversity on the different boulder classes?
Predation – here the hypothesis parallels what was described
for keystone predators. The intensity of competition among
prey species is reduced, thus permitting a greater number of
species to persist. This is especially true when the predators
follow a type III functional response. If predation is prey
frequency dependent, then prey diversity can increase.
When we talk about diversity, we have to careful to identify
the scale we are talking about.
Within a single community, there is a diversity of species
present. At the local level, we call this α diversity.
There is a second, larger level in which we identify the
variation in the species lists of communities in different
habitats within a region. This is β diversity.
Finally, there is a third level that considers the total species
diversity over the whole region. This level is called γ
diversity.
γ diversity = average α diversity x β diversity
This relationship can be viewed in a slightly different way.
γ diversity indicates the entire species pool available in a
region.
Not all species from this pool are present in each
community. Instead, there is a process called species sorting
that separates the total pool into groups of species that can
successfully coexist in single communities.
One of the best (and few) studies to investigate sorting was
work done by Paul Keddy of the University of Ottawa. He
planted seeds of 20 different wetland species (his regional
pool) into 120 different wetland areas, among which
physical and chemical conditions differed.
He then followed what happened for 5 years…
One species failed to germinate in any of the 120 sites.
5 others didn’t persist in any community for the full 5 years.
That left 14 species to be sorted by the differences in physical
and chemical conditions of the individual plots. Ricklefs
plotted the results in terms
of the average number of
species in individual plots,
separated into those from
plots with fertile soil and
those from infertile soil.
fertile soil