Thesis #714

Many animal and plant species occur in
metapopulations that are assemblages of spatially
delimited local populations coupled by some degree
of migration. The occupation of habitat patches may
change over time due to local extinction and
colonisation. Migrants colonise empty habitat
patches, which often leads to founder effects. They
may also invade existing populations, and thereby
increase local genetic diversity. Besides gene flow in
the focal species, migrants may evade parasites or
parasites may co-migrate with their hosts. Often,
migrants are not a random subset of their population
of origin, and populations may differ in their
contribution of migrants. Due to the evolutionary and
ecological significance of migrants, it is important to
know their number and populations of origin.
Parasites may drive the evolution within host
populations. But they also influence the success of
migrants and thereby gene flow between populations.
I studied ecological and evolutionary
dynamics in Daphnia metapopulations. Three
interacting species of Daphnia C namely D. magna,
D. longispina and D. pulex C occur sympatrically
along the coast of southwest Finland. They live in
ephemeral freshwater rock pools of various size and
reproduce asexually during most of the summer. The
sexually produced migration stage, the so-called
ephippium, is essential to survive harsh
environmental conditions such as desiccation during
summer or the freezing during winter. There exist
two different hypotheses on the origin of migrants in
this metapopulation. One hypothesis assumes a
Levins' type metapopulation, with no differences
between the patches, while the other suggests a
mainland-island model, where long-lived populations
in large patches are the source of migrants. In a first
step, I quantified the ephippium production of
populations in various sized natural rock pools and in
containers under outdoor conditions. Populations in
larger habitats produced more ephippia but the
increase was much smaller than the increase in
habitat size, and the numerical dominant populations
in small rock pools produced a substantial number of
ephippia.
In the next chapter, I show that desiccation,
which is a common phenomenon in the natural rock
pools, is not detrimental for the populations. This
year's ephippia are sufficient to survive a desiccation
event and an ephippium bank from previous years is
not required. I developed a mathematical model to
predict desiccation for more than 500 individual
pools over 25 years. During warm and dry periods,
evaporation is high and especially shallow pools with
a small surface area and vegetation tend to dry up.
Mevertheless, also these pools with a high risk of
desiccation are often inhabited by Daphnia
populations. Populations in these ephemeral pools are
usually short-lived, but ephippia are especially
exposed to passive dispersal by wind or birds in the
sediments of desiccated pools. I showed that
populations in small pools (less than about 300 l
volume) produced about 50 % of all ephippia.
Rowever, 90 % of all exposed ephippia originated
from these populations. Exposition of ephippia on dry
sediments is practically non-existing in pools larger
than 1000 l. This analysis suggests that populations in
small ephemeral pools are most relevant for the
metapopulation dynamics.
Consistent with the predictions and the
functional understanding of the production of
migrants, I found increased colonisation rates after
warm and dry summers. The weather in southwest
Finland changed in accordance to global climate
change predictions, and this led to increased
dynamics in the metapopulations of the three
Daphnia species. It is the first time that an influence
of climate change on metapopulation dynamics has
been shown. Furthermore, I also found changes in the
whole metacommunity composition, as the three
species reacted differently to climate change.
In the final chapters, I did not focus on
migrants themselves, but investigated proximate
effects of migration. All ephippia can migrate, but the
successful invasion and establishment depends on
fitness components of the hatchlings. For example,
migrants infected with the microsporidium
1ctosporea bayeri are less successful than uninfected
migrants. I was interested in further correlations
between the fitness of a host and its natural infection
status and compared the cost of resistance hypothesis
with the inbreeding-infection hypothesis. I
experimentally confirmed that the naturally observed
infection status has a genetic basis. However, I did
not find a difference in competitive abilities between
naturally uninfected and cured but former infected
genotypes. This suggests that resistance genes
segregate independently of other fitness associated
genes in this system.
A consequence of migration and dispersal is
either the establishment of low-diversity and
potentially parasite-free populations in newly
colonized habitat patches or the introduction of new
host genotypes into already existing populations,
which increases local genetic diversity. Parasites may
either co-migrate with their hosts or arrive
independently in D. magna populations. I thus
compared the epidemiology of O. bayeri in host
populations of low and high genetic diversity.
Following parasite prevalence over two years, I
showed that the parasite spread less successful in host
populations of higher genetic diversity. In the long-
term, this may influence coevolution and hamper
local adaptation of the parasite.