Transcript Seed and pollen dispersal

Seed and pollen dispersal
Remember from the first lecture that seeds may be dispersed
on wind, by water, and by animals. Animal-borne seeds may
be carried externally or may be transported until defecated
from the gut.
The basic spatial pattern for dispersal is very similar for all
mechanisms of dispersal. All show a peak in abundance of
dispersed propagules at a short distance from the source plant,
then a decline with increasing distance. The rate of decline
differs, but all distributions are skewed, with a more-or-less
extended tail.
Think of a wind-dispersed seed as one example. Seeds are
broken free of the parent (abscised) by a sufficiently strong
velocity (a gust possibly) of wind.
The force required to break the seed free explains why the
peak of the distribution occurs at some small distance from
the parent.
The remainder of the distribution is dependent on what type
and size of dispersal accessory structure the seed carries.
Some seeds have no dispersal accessory. They are sometimes
called ‘ploppers’, and the number dispersed drops very
rapidly with distance.
Other seeds (particularly a number of common weeds) have
‘parachutes’ that make possible relatively long distance
dispersal.
A dandelion, in case you
didn’t recognize it.
The distance that animals disperse seeds may be determined
by how far the animal moves during gut passage time, or how
far it moves before its behaviour (scratching, rubbing against
something) releases the seed.
Yet these radically different kinds of dispersed propagules
have similar curves of the number of propagules dispersed
differing distances.
The figure on the left represents microsatellite markers that
identify distances of saplings of bur oaks from the parents. It
may not estimate acorn dispersal accurately. The figure on the
right shows dispersal distances for propagules of prairie
fugitive species. Note that here peaks are not at 0 distance.
Pollen movement follows a similar pattern.
This figure shows pollen dispersal distance for a number of
species. Note that the x-axis is log transformed – there would
be a more-or-less long tail on each of these distributions, and
that many peak meters from the parent. However, also note
that pollen can travel much further, on average, than seeds.
Pollen may be moved by insect or bird pollinators. Then
dispersal distance is determined by the foraging behaviour of
the pollinator, and may be much more local.
The peaked distribution (filled circles) are interplant flight
distances for the pollinator. The open symbols, a much flatter
distribution, are the distribution of interplant distances and
that for realized pollen dispersal. The plant, Asclepias
exultata, uses pollinia (packages of pollen grains).
The picture is of an flower of Asclepias sullivantii. The
‘lumps’ in the middle are pollinia, each about 3mm long. A
pollinator moves whole pollinia from this flower onto
another one.
Pollinia on the leg of a carpenter bee
Copyright © 2004 Beatriz Moisset
What is important about all these distributions is the relatively
long tails – the skewness of the distributions. Long distance
movements are what produce new population isolates, what
produce gene flow in continuous populations to reduce
subdivision, and, in general, what determines the amount of
genetic structure in populations.
Without reference to mechanism, we can also look at the
distance between mating parent plants (determined by
molecular markers that reveal parentage). These data are for a
dioecious lily.
At the opposite end of pollen dispersal is the exchange of
pollen among flowers of the same plant (or genet). This kind
of pollen exchange is called geitonogamy.
It can occur in species that are self-fertile, but it can also ‘gum
up’ receptive surfaces if the species is self-incompatible.
The amount of geitonogamous pollen transfer apparently
depends in part on the number/density of flowers on a plant/in
an area.
In the musk mallow, Malva moschata,
selfing rate varied with the number of
flowers on plants. More selfing occurs
on plants with larger numbers of flowers
(since bees probably spend longer moving
among flowers on those plants).
One of the best examples that show pollen and seed dispersal
have separate impacts is the classic example of the genetics of
mine spoil grass, Agrostis tenuis, that grows on and near
Welsh lead mines.
The mines are relatively recent (<100 years old).
The grass grows both on normal soil and on mine tailings that
have high levels of toxic heavy metals.
Consider the course of evolution of tolerance: Mine spoils
were originally bare soil. Only seed dispersal could initiate a
population of Agrostis. There must have been very severe
selection for tolerance on seedlings and plants growing from
those seeds.
Tolerant plants produce pollen carrying the tolerance gene(s).
There is a fairly stable prevailing wind at the mine site.
Seeds are blown onto the site, but pollen is dispersed across
the site and beyond. Tolerant and non-tolerant pollen is
equally successful at producing seeds.
Tolerance costs energy, so that resistant plants are inferior
competitors to non-tolerant ones when growing without heavy
metals in the soil. Tolerance, and the energy expenditure to
achieve it, is an evolved trade-off against growth rate, etc. in
the absence of tolerance.
The continued presence of both types of grass depends on the
presence of mine spoil soils.
The pattern of tolerance is observed not just in Agrostis
tenuis, but also in another grass present on the spoils,
Anthoxanthum odoratum.
As a question in evolution, this example explains how
diversifying selection can occur.
In a more general context, we can compare the impact of
pollen and seed dispersal:
Only seeds can found new populations
Pollen can only spread genes through existing populations
Track the contributions of pollen and seeds by looking at
nuclear markers (found in both seed and pollen) vs.
maternally inherited genes (e.g. genes in chloroplasts and/or
mitochondria, but not found in pollen). Dispersal distances
differ between wind and animal dispersal…
The estimated ratio of contribution to plant genotypes from
pollen and seed was higher in the wind-pollinated, highly
outcrossed tree Quercus petraea (sessile oak) than in the
selfing grass Hordeum spontaneum (a wild barley). The
ratios were 196 vs. 4.
Now we can consider one explanation of how selfing can be
selectively advantageous:
Outcrossing plants growing at low density may fail to receive
enough pollen to breed.
That can lead to selection for self-fertilization to assure
reproduction. Fitness could be very low for a plant that can
only outcross when it is isolated or its density is low.
This is one explanation for the breakdown in outcrossing
systems at the edges of habitats and at the edges of species’
ranges.
This brings us back to considering isolation by distance…
The probability of two specific individuals mating decreases
as the distance between them increases.
Combined with limited dispersal of seeds and pollen,
neighbours tend to be related (their alleles originate from a
common ancestor)
This leads to a genetic structure within a population– which
we can look at as the likelihood that two randomly selected
alleles have come from the same ancestor.
Some standard for that likelihood is used to measure
neighborhood size - the size of an unstructured population
that would produce the same probability of identity by
descent.
If we assume seeds and pollen disperse equally in all
directions around the parent, and the standard deviation of
dispersal distance between parent and offspring is 2, then the
genetic neighborhood size is:
A = 42
And we have the information to calculate the effective
population size – it is the number of plants in the genetic
neighborhood size:
Ne = NA
Where this N is the density (N.B. not number) of
reproductive plants within the area.
For the same reason that selfing tends to increase where floral
density is higher, the neighborhood area tends to become
smaller:
Bees (or other pollinators) don’t have to travel as far to move
from one nectar and pollen source to the next…
One example is a prairie legume, Chamaecrista fasciculata
(Partridge pea).
It has a high outcrossing rate (0.8), but a
remarkably small neighborhood size (2.4m).
The population size within the neighborhood
is only about 100 plants.
Given knowledge of seed dispersal, as well
as pollen movement, it was possible to
partition the contributions:
A partition of
gene flow
Seed dispersal contributes little, but pollen dispersal is a major
contributor to gene flow.
The largest factor was distance-dependent progeny fitness.
Fitness increases from selfing to outcrossing within a
neighborhood (likely including some inbreeding depression,
but still 2x selfing) to outcrossing involving parents more than
one neighborhood radius apart.
The Founder Effect
Founders of a new population likely carry only a small sample
of alleles present in the source.
Thus each new population will have differing allele
frequencies.
Bottlenecks have a similar effect: reduction of a population to
a small size that then recovers from surviving founders.
Both founder and bottleneck effects reduce genetic diversity.
The small sample of the founder population may not be
representative of the allele frequencies present in the larger
source population, and rare alleles may be missed.
At the population or species level, the loss of genetic diversity
due to the founder effect should be small. Why?
Answer: Alleles lost in any group as a founder effect are
likely to have been rare in the source population.
Those rare alleles contribute little to He (gene diversity,
calculated from the frequencies of alleles at a locus as He =
1 - pi2) they make negligible contribution – with small pi,
their pi2 will be small.
Alleles with intermediate or high frequency are likely to
persist in newly founded populations.
That is the case with the example of Asclepias exaltata
(poke milkweed). A northern population has lost 19
uncommon alleles still present in populations in the
southern Appalachians.
This leads to a broader principle: the Wahlund effect
Given lower genetic diversity in small subpopulations, due to
founder effects, bottlenecks, or selection adapting the
subpopulation to its local environment, there frequently are
higher allele frequencies for the most common allele. What
will the Hardy-Weinberg equilibrium frequencies look like in
each subpopulation?
Assume p = 0.8 and q = 0.2
p2 = 0.64 - frequency of the common allele homozygote
2pq = 0.32 – frequency of heterozygotes
q2 = 0.04 – frequency or ‘rare’ homozygotes
Overall frequency of homozygotes is 0.68
If, in another subpopulation, these p and q are reversed, there
will still be a higher fraction of homozygotes.
Averaged over the whole population, we would expect the
usual 50% homozygotes, since total population p = q = 0.5.
However, given the effects of sampling (the founder effect,
bottlenecks) and/or local selection, when we count
homozygotes across the population we find an excess.
This excess is what is called the Wahlund effect.
There is another general principle, called Baker’s rule, after
Herbert Baker.
Baker’s Rule:
Successful founding populations should often be self-fertile.
Why?
Founders have few neighbors.
Neighbors are likely to be close relatives.
Self-incompatible and same-sex dioecious colonists are at a
disadvantage.
What are the patterns of genetic diversity in plant
populations?
The patterns we see in genetic diversity are functions of
effective population size and deme isolation.
Genetic drift causes a loss of genetic variation in isolated
subpopulations and differentiation among them.
Inbreeding will accelerate subpopulation differentiation, since
migration rates of alleles among subpopulations will be very
low. Inbreeding also reduces effective population size and
increases homozygosity.
Thus there should be lower genetic variation in inbred demes
phlox – sweet william
oenothera – evening primrose
lolium – rye grass
plantago – plantain
mimulus – monkey flower
lilium – lily
arenaria - sandwort
That result is apparent when we consider species within
genera that include both inbreeding and outcrossing species.
Note that the Y-axis is Hs, the within population diversity
(heterozygosity).
Since inbreeding subpopulations tend to be isolated, they may
also be subject to strong adaptive selection to local conditions,
and that can also increase differences.
In inbreeding populations, selection can choose the best
adapted homozygote (or at least more likely to be a
homozygote in these populations).
Inbreeding perpetuates that homozygote into subsequent
generations.
If outcrossed, the successful phenotype may not be
perpetuated; instead, a variety of phenotypes will be produced,
only a fraction of which will be successful.
The table below shows mean proportion of genetic variation at
allozyme loci for different breeding systems:
Patterns in this table??
An aside: the gene statistics that have arisen in this lecture:
He – gene diversity (or expected heterozygote frequency) –
H e  1   pi2
i
HS – subpopulation gene diversity – the equivalent of He for
an isolated group, with an identical formula summing only
over alleles found within the group.
HT – total population gene diversity, obtained by summing
over the set of demes forming the total population. If there
are founder effects or drift occurring within individual
demes, then HT will be larger than HS.
FST – the between population proportion of gene diversity
(or, in different words, the extent to which individuals in
subpopulations are more similar than in the total set of
HT  H S
populations) F 
ST
HT