The marsh fritillary
butterfly is one of the nine British fritillary species. It is a
reddish brown insect with cream, yellow and dark markings and has
spiny larvae that are dark in colour. It is primarily a wetland
species and has a very restricted range throughout Europe that in
Wales is associated with marshy grassland habitat containing the
larval food-plant devil's-bit scabious Succisa pratensis.
Individual site populations are prone to wide fluctuations in size;
there are tales from Ireland in the nineteenth century of massive
'outbreaks' of caterpillars causing villagers to barricade their
homes and rake up huge piles for burning (Thomas and Lewington
1991). The butterfly occurs in groups of populations known as
metapopulations, where each population is no more than about 5 km
from another population. Occasional interchange of adults between
individual populations reduces the occurrence of localised
extinctions. Up to the 1930s it is very likely that the Rhos
Llawrcwrt fritillaries would have been part of a large
metapopulation.
In west Wales, adult
emergence usually starts in the 3rd week in May and is finished by
the 3rd week in June. The butterflies are very weak fliers and even
mature hedges can act as barriers to movement. They are
particularly susceptible to poor weather during the flight period.
Wind and rain can reduce the chance of successful pairing and egg
laying. The existence of natural shelter in the breeding habitat
can be very important and, in years when the weather is poor and
the population is low, may be a critical factor in successful
breeding.
Adults are rarely on
the wing after the 1st week in July. Eggs are laid in rows on the
underside of leaves of the food-plant and hatch after 3 weeks. The
caterpillars immediately spin a dense silken web within which they
feed until all leaves are devoured. They then move en masse until
the next plant is located and repeat the process again. This
continues until late August when the caterpillars spin a dense silk
web deep in the vegetation, often in the base of grass tussocks.
They spend the winter months in this web and reappear on the first
warm days of spring, often as early as late February. As air
temperatures are low at this time of year and the caterpillars must
raise their body temperature to feed, they cluster together above
the vegetation, sometimes on top of webs, to form a black mass of
bodies that absorbs the heat of the sun. Feeding in webs continues
until eventually they start to wander away from the webs and feed
as individuals until pupation, which will normally start from
mid-April. Optimum habitat structure may be described as a mix of
both short sward vegetation and tussocky vegetation to provide
areas for both basking and shelter. The desired sward height at the
end of the grazing season is usually quoted as being between 8 and
25cm. This structure can only be maintained by light summer grazing
with cattle or ponies. Stocking rates need to be in the order of
0.2-0.4 lu/ha/annum (livestock units per hectare per annum),
although this will need to vary from year to year to take account
of vegetation growth.
Apart from the usual
range of predators, including ground beetles and spiders, the marsh
fritillary is parasitised by tiny parasitic wasps. In Britain there
are two species, but in Wales the wasp Apantiles bignelli is the
main parasite. This wasp injects eggs into the caterpillars; wasp
grubs emerge from these eggs and feed on the butterfly caterpillars
until they are ready to spin their own cocoons and finally emerge
as adult wasps. Up to 70 wasps may emerge from a single caterpillar
and each generation of butterfly caterpillars may be host to up to
three generations of wasps. The second generation of wasp grubs
overwinters in the butterfly caterpillars. The wasp can have a
devastating effect on fritillary populations, killing up to 75% of
caterpillars in some years.
The relationship
between the parasite and fluctuations in the butterfly population
is not fully understood, but a reasonable model is based on a
multi-year cycle where the wasp population expands to a point that
causes a crash in the butterfly population which in turn causes a
crash in the parasite population. This host-parasite cycle is
affected by many other parameters, including the weather. Cool,
sunny springs favour the butterfly, the caterpillars of which warm
up quickly by communal basking. Wasp pupae develop much more slowly
in lower temperatures, and wasps emerge too late to parasitise the
butterfly caterpillars before they pupate. A cycle of fixed length
between peaks in butterfly population is therefore possible but
unlikely. As is the case with most host-parasite relationships, the
parasite alone is unlikely ever to be the cause of long-term
decline or extinction in the host.
For a site manager,
the challenge is to distinguish responses in the butterfly
population to 'natural factors', such as the parasite or climate,
from factors over which the site manager is able to exert some
control, such as habitat structure and quality.
It was not until
1983 that any data on the status of the butterfly at Rhos Llawrcwrt
was collected. Adrian Fowles (CCW entomologist) surveyed all known
marsh fritillary sites in Ceredigion, using direct counts for all
adults at small sites and counts along a series of parallel
transects over suitable habitat at larger sites (Fowles 1983).
Fowles recorded 413 adults at Rhos Llawrcwrt, by far the highest
count in the survey that year, representing about 75% of the total
count for all sites. Extrapolation of the count data indicated that
a conservative estimate of total marsh fritillaries flying on the
site on the count day would be in the order of 700. Assuming that
was a peak figure for the entire emergence period, it was suggested
that in excess of 2,000 adults emerged at Rhos Llawrcwrt in 1983
(Fowles 1983). However, the Ceredigion-wide survey also confirmed
that Rhos Llawrcwrt was an isolated site and not part of a larger
metapopulation. It was therefore very vulnerable.
In 1984, Fowles
established a Pollard Walk to collect data on adult butterflies,
following the methodology described in the National Butterfly
Monitoring Scheme (Pollard 1977). This was based on a transect
route in the western block of the NNR, approximately 1.7 km long,
that was sampled throughout the flight period. From 1985, this
route was walked once each week during the flight period, but only
with climatic parameters in which the butterflies would be active.
All adult butterflies within a defined distance from the recorder
were counted. The total number of butterflies counted for all weeks
was than recorded as an annual adult index. Fowles also set out a
series of 53 100 m2 transects to collect data on larval webs. These
web transects were located so as to representatively sample all
potential breeding habitat available in the western block at the
time. The transects were recorded during early September, and the
total number of larval webs observed was recorded as an annual
larval web index.