Individual study: Creating woodland glades for the conservation of the woodland brown butterfly Lopinga achine in Östergötlands, Sweden
Bergman K.O. (2001) Population dynamics and the importance of habitat management for conservation of the butterfly Lopinga achine. Journal of Applied Ecology, 38, 1303-1313
Like many butterfly species in Western Europe, the woodland brown Lopinga achine, has disappeared from many former localities in recent decades. The primary reasons for its decline are considered to be habitat loss and degradation through a decline in traditional woodland management; the woodland brown is reliant on woodlands with open glades where its larval food plant, soft-leaved sedge Carex montana, grows. Five populations of woodland brown in the province of Östergötland, southern Sweden, disappeared in the 1960s and 1970s, probably due to the sites becoming too overgrown, even though the larval food plant was still present. As a prerequisite to initiating a species recovery programme, a greater understanding of the butterfly's ecology was required, including a better idea of optimum tree and shrub canopy cover, and how the creation of new glades in neglected woodlands would effect populations.
Study area: The study area was located 30 km south of the city of Linköping, southern Sweden, in an area of 50 km² encompassing five deciduous woodland sites supporting woodland brown populations. A site was defined as a woodland patch surrounded by at least 20 m of unsuitable habitat (in most cases open fields or spruce Picea abies plantations). The sites where characterised by a partly open tree canopy dominated by oak Quercus robur interspersed with a hazel Corylus avellana understorey.
Population sizes: Peak adult population sizes were estimated from 1989 to 1997 using the Jolly-Seber mark-release-recapture method (MMR). Each captured adult was sexed and marked (using a permanent marker pen) on the underside of one hind wing to allow individual recognition, and released at the point of capture. The sites were visited regulrly during the flight period (mid-June to mid-July).
This data indicated that the peak flight period commenced two days after appearance of the first females. From 1992 to 1998, the start of this peak period was estimated by checking the sites daily for the first appearance of females. In 1992-97 the peak population sizes where estimated by MRR 3-4 days in a row; in 1998 they were estimated based on the numbers seen when walking along a route encompassing all favoured glades. Looking at the results from the 1997 MRR survey, there appeared to be a close correlation between the two survey techniques.
The total number of adults per brood was calculated on the basis of an earlier study that indicated that 35% of all individuals fly at the peak period.
Management: In 1992-95, vegetation was cleared to create new glade edges at six unmanaged sites where the risk of L.achine extirpation was considered high because few open glades remained. New irregularly shaped glades 10-30 m long were cut. Where possible the longest side had a southwest-northeast orientation for maximum exposure to the sun.
In spring 1992, 20 glades at the first site were cut and population size estimated at the peak flight period. At the other sites, adult population size was estimated prior to creating five or six glades. Four sites were managed in 1992, one in 1993 and one in 1995. At the same time, at three to six reference sites where no glades were created, and traditional management had also been abandoned, population estimates were also made.
Sedge abundance: Changes in abundance of soft-leaved sedge Carex montana, was recorded in 84 permanent 1 m² quadrats at the six managed sites at the same intervals that butterflies were surveyed. Quadrats were located in four different areas (21 in each):
i) edges of created glades where established C.montana plants were present;
ii) edges of created glades where C.montana plants was absent;
iii) closed woodland where established C.montana plants were present;
iv) closed woodland where C.montana was absent.
Each year in July, a 1m² quadrat, was laid over the plot and the percentage coverage of C.montana was estimated.
Egg production and mortality: Twenty female butterflies were captive-reared and used to estimate number of eggs produced. Egg production was subsequently estimated using the daily survival of 117 females from MRR surveys.
Under natural conditions, females drop their eggs onto the ground, one by one. Egg survival was estimated in the field using 176 newly laid eggs from 13 freshly caught females in two populations, the eggs being obtained in the laboratory. These eggs were placed in three different glade environments where ovipositing females were observed as follows:
i) 58 in a sun exposed position, 3-6 m from the edge of the glade;
ii) 61 along the edge of the glade, 0.2-0.5 m under the tree and bush canopy;
iii) 57 in the shade under the tree/bush canopy, 2-5 m from the edge.
Eggs were placed singly on hazel leaves, secured to the ground/litter layer with a needle. Eggs were checked daily for predation and subsequent hatching.
Larval mortality: Mortality was estimated for 280 larvae, 140 in 1994-95 and 140 in 1995-96: Twenty newly hatched larvae were placed in each of seven 1.2 x 1.2 x 0.7 m cages in edge habitats not exposed to the sun or within the shade habitat. Cages were open at the top but to prevent the larvae escaping there was a 10 cm wide lip, the upper surface of which was coated with trapping adhesive (Insect trap coating: Tanglefoot Co.). No larvae were found stuck to the adhesive in either study year. At the time of pupation, the cages were covered to prevent predation.
Pupal mortality: Pupal mortality was estimated using 43 pupae (37 laboratory reared, six found in the field) in edge habitats. The reared larvae which had pupated on the net cage lids, were removed by cutting the net, attached to sticks and placed in the field at ground level to mimic natural conditions. The six wild pupae were left in situ where they had been found. Pupae were checked daily.
Tree and shrub cover, and rate of canopy closure: Using aerial photographs taken in 1941, 1962-63, 1984 and 1997, the area not covered by trees and shrubs was marked and the extent estimated. The rate of canopy closure in woodlands not affected by domestic animal grazing or forestry was calculated. Using this information, canopy cover was predicted for the years 2017 and 2037 i.e. 20 year intervals, assuming that present management of these sites remained the same.
Population dynamics and population size: Management of the four overgrown sites in which the first glades were created in 1992 had very positive results with an average population increase of 97% between 1992 and 1997. In comparison in these years, populations at the six reference sites (no management) decreased by 25%. The population of the site managed in 1993 also increased substantially, by 93% from 1993 to 1997, whilst the three reference sites increased by 19%. For the site managed in 1995 the results were very different with a 27% decrease in population and a 9% decrease at the reference sites. The cause of these declines is unknown, but may have been due to a stochastic event and/or that two years was too short a time period for the population to adjust to the newly created habitat.
Most populations in the study area were small: 50 of the 79 populations comprised less than 500 individuals, five were larger than 1,500 individuals and the largest, between 4,000 to 4,500 butterflies. Populations fluctuated synchronously between years, some years being better than others, e.g. all study populations increased from 1991 to 1992 and all decreased sharply (32-83%) from 1992 to 1993. Generally, 1995 and 1996 appeared to be good years.
Long-term population dynamics and occurrence were closely correlated with tree and bush cover. Populations occurred only at sites with greater than or equal to 60% canopy cover but population density decreased sharply where cover exceeded 90%. There was a peak in population density at an optimum canopy cover of 70% to 75% and a decline as succession continued.
Sedge abundance: Cover of the host-plant C.montana increased significantly at edges of new glades and decreased in closed unmanaged woods. Successful restoration probably requires the presence of C. montana along edges of new glades from the onset of management because it was slow to colonize plots where it was initially absent. Of the 42 plots that lacked C.montana at the start of the experiment, only one new plant appeared in one plot at the edge of a created glade, and one each in two plots in closed unmanaged areas.
Egg to adult survival: Survival from egg to adult was highest at the edges of glades (2.3 adults per female) and lowest in the wood shade (0.7 adults) and sun exposed (0.6 adults). The life table (see Table 1, attached) shows that a population of L.achine can only be maintained in glade edge habitat, and values for the shade and sun-exposed environments may infact be over-estimates, as calculations of larval and pupal mortality were based on estimates for the preferred edge habitat only.
Egg production: The potential number of eggs per female was 140 based on the average of 20 individuals. The realized production of eggs based on the survival of 117 females in the field was estimated to be 77 eggs.
Egg mortality: The mortality of eggs was lowest at the edges of glades (52% of 61 eggs) and higher in the open areas (86% of 58 eggs) and in the shade (88% of 57 eggs). Causes of mortality were also different in each of these three habitats. The proportion dying from dessication was highest in open areas exposed to the sun (48%) compared with glade edges (9%) and shaded areas (0%). However, predation (predator unknown) was highest in shaded areas (98% of eggs), and accounted for 84% of mortality along edges and 48% in the open areas.
Larval mortality: Of the 280 larvae studied in the open cages, only 11 of 140 survived to imago in 1994-95 and only 16 of 140 in 1995-96. This should be viewed bearing in mind that the larval period lasts 10 months which constitutes 80% of the life span of L.achine, but the mortality during the larval period is greater than any other life stage, estimated at 90%.
Pupal mortality: Pupal mortality was estimated at 36% and thought primarily due to predation as pupae either disappeared or remnants of the pupal case were found. Three pupae darkened but never hatched.
Rate of canopy cover: The annual rate of canopy closure at unmanaged sites decreased linearly with tree and bush cover, approximately 1% closure at 60% cover and 0.3% at 80% cover, making it possible to predict the impact of habitat changes for L.achine. 86% of sites in southern Sweden occupied by L.achine are presently unmanaged and it is predicted that many populations will become extirpated within 20-40 years if this situation continues.
Conclusions: The results of this study suggest that the creation of new glades in overgrown woodland habitat is beneficial to dwindling populations of L.achine. Recovery programmes for L.achine should therefore be aimed at producing the required successional conditions by creating suitable glade habitat whilst maintaining a woodland canopy cover of around 65% to 75%. Such management should also benefit C.montana, the larval food plant, which shows a reduction in abundance at high levels of shading as canopy cover increases, as found in unmanaged woodlands.
Note: If using or referring to this published study, please read and quote the original paper. Please do not quote as a conservationevidence.com case as this is for previously unpublished work only.