Journal of Animal

Ecology

2007

76

,



12241230

2007 The Authors.
Journal compilation
2007 British
Ecological Society

Blackwell Publishing Ltd

What is the required minimum landscape size for dispersal
studies?

MARKUS FRANZN and SVEN G. NILSSON

Department of Animal Ecology, Ecology Building, Lund University, SE-223 62 Lund, Sweden

Summary
1.

Among small animals dispersal parameters are mainly obtained by traditional
methods using population studies of marked individuals. Dispersal studies may
underestimate the rate and distance of dispersal, and be biased because of aggregated
habitat patches and a small study area. The probability of observing long distance
dispersal events decreases with distance travelled by the organisms. In this study a new
approach is presented to solve this methodological problem.

2.

An extensive markreleaserecapture programme was performed in an area of
81 km

2

in southern Sweden. To estimate the required size of the study area for adequate
dispersal measures we examined the effect of study area size on dispersal distance using
empirical data and a repeated subsampling procedure. In 2003 and 2004, two species of
diurnal burnet moths (Zygaenidae) were studied to explore dispersal patterns.

3.

The longest conrmed dispersal distance was 5600 m and in total 100 dispersal
events were found between habitat patches for the two species. The estimated dispersal
distance was strongly affected by the size of the study area and the number of marked
individuals. For areas less than 10 km

2

most of the dispersal events were undetected.
Realistic estimates of dispersal distance require a study area of at least 50 km

2

.

4.

To obtain adequate measures of dispersal, the marked population should be large,
preferably over 500 recaptured individuals. This result was evident for the mean moved
distance, mean dispersal distance and maximum dispersal distance.

5.

In general, traditional dispersal studies are performed in small study areas and based
on few individuals and should therefore be interpreted with care. Adequate dispersal
measures for insects obtained by radio-tracking and genetic estimates (gene ow) is still
a challenge for the future.

Key-words

: dispersal distance, Lepidoptera, markreleaserecapture, movement,
Zygaenidae.

Journal of Animal Ecology

(2007)

76

, 12241230
doi: 10.1111/j.1365-2656.2007.01285.x

Introduction

Accurate estimates of distances travelled by different
species is central to metapopulation ecology, gene
ow estimates, population dynamics, species survival
in fragmented landscapes, and species conservation
actions, e.g. reserve planning (Hanski & Gaggiotti
2004). Yet, for most small animals, dispersal is one of
the most difcult population parameters to measure
(Nathan 2005). Direct traditional estimates, e.g. based
on markreleaserecapture (MRR) techniques, often
result in biased dispersal data reecting the study area
size (Barrowclough 1978; Koenig, van Vuren & Hooge
1996). Schneider (2003) examined different MRR
studies performed with butteries and concluded
that the measured mean dispersal distance increased
with study area size. Butteries have become model
organisms in metapopulation studies (Hanski 1994;
Hanski, Kuussaari & Nieminen 1994), but remarkably
few studies have critically examined the effects of
study area size on dispersal parameters (cf. Englund
& Hambck 2007). Thus, even if dispersal distances
are substantially underestimated in many studies
(Barrowclough 1978; Shreeve 1992; Wilson & Thomas
2002), it is not obvious what study area size is required
for a realistic measure of dispersal distances.

Correspondence: Markus Franzn, Department of Animal
Ecology, Ecology Building, Lund University, SE-223 62
Lund, Sweden. Tel.: +46 46 222 38 20. Fax: +46 46 222 47 16.
E-mail: Markus.Franzen@zooekol.lu.se
1225

Required area to
measure dispersal

2007 The Authors.
Journal compilation
2007 British
Ecological Society,

Journal of Animal
Ecology

,

76

,



12241230

This study presents a new approach to obtain
accurate estimates of dispersal distances and applies
the new method to data from two species of burnet
moths (Zygaenidae). Based on our data we give
recommendations on required study area size and the
number of recaptured individuals required for accurate
estimates of dispersal parameters for species with
similar dispersal patterns.

Materials and methods

s1inx sirci rs

The burnets

Zygaena viciae

(D&S.) and

Z. lonicerae

(Scheven) have similar ecology and a single annual
generation in Sweden, with a main ight period in
July. These two species are dependent on ower-rich
seminatural grasslands and occurs locally in most of
Europe (Naumann, Tarmann & Tremewan 1999). Both
species have recently declined in many west European
countries (Visnen & Somerma 1993; Naumann

et al

.
1999). In the study area mating takes place on nectar
rich owers, mainly on the herb

Knautia arvensis

(L.)
(Dipsacaceae). When approached the moths remain
exposed on the owers, making them ideal for studies
of dispersal.

s1inx :r:

The study area, covering 81 km

2

(Fig. 1), was situated
in Stenbrohults parish in southern Sweden (56

37

N,
14

11

E). The area borders the large Lake Mckeln in
the west, and forest-dominated land to the south, east
and north. The landscape in this region is mixed,
although forest dominates (> 75%).

n:ni 1:1 i:1cnrs

All open grassland patches, about 5% of the study
area with seminatural grasslands covering 35%, were
mapped in 2003. Grassland patches occupied by the
study species in 2003 and/or 2004 were dened as
habitat patches and each patch considered separate if
the distance to the nearest patch was 100 m or more.
Patches separated by tall tree stands, an additional
barrier to dispersal, were also regarded as separate
habitat patches if the distance between them was 75 m
or more. The two burnet species were recorded from 68
habitat patches with a mean size of 12 ha (range 006
81) (Fig. 1). At 38 patches more than 10 individuals
per patch were marked. These 38 patches ranged in size
from 006 to 50 ha (mean 139) and were all situated
on abandoned grasslands (previously pastures) or hay
meadows with late harvest. The amount of grassland
occupied by burnet moths was low, only 06% of the
study area (Fig. 2). The proportion of the area surveyed
and the proportion occupied by burnet moths
reached a maximum 15 km from the study area centre
and decreased towards the edges of the study area
(Fig. 2).
Fig. 1. The study area in Stenbohult, southern Sweden.
Habitat patches occupied by burnet moths are lled in black
and other grassland patches unlled areas.
Fig. 2. (a) The total area (), the area of grassland
habitats (), and the area occupied by burnet moths
(), in increasing distances from the study area centre.
The area in log 10 transformed hectares. (b) The proportion
of the total study area that was surveyed for burnet moths
(i.e. the proportion of grassland habitats) (), the
proportion of the study area that was occupied by burnets
moths (), in increasing distances from the study area
centre.
1226

M. Franzn &
S. G. Nilsson

2007 The Authors.
Journal compilation
2007 British
Ecological Society,

Journal of Animal
Ecology

,

76

,



12241230

x:xrir:srrc:i1ir

An extensive MRR programme for burnet moths was
carried out 26 June31 July 2003 and 23 June9 August
2004. The MRR programme started when the rst
individual was observed in the study area and continued
throughout the ight period of the species. We dene
dispersal as a movement between habitat patches by
one individual. It was impossible to separate natal and
breeding dispersal (cf. Paradis

et al

. 1998). At each
marking occasion the date and the species name, sex,
activity and habitat patch of capture were recorded. All
patches were visited every third day during ideal
weather conditions and all encountered individuals
were captured with a buttery net. Every patch was
searched with equal frequency and intensity. Each
individual caught was marked with a ne-point per-
manent ink pen. Unique marking codes were used at
each patch and occasion of marking. Recaptured
individuals that had moved between patches were
subsequently individually marked according to Kreusel
(1999). Using this simplied marking method we could
extend the study area and increase the possibilities of
nding long distance movements between patches.
With this method it is impossible to track each single
individual but each marked animal could be traced to
patch and date. This information is crucial when
studying dispersal patterns, as opposed to movements
within a habitat patch where it is necessary to trace
each animal individually.
Recaptures occurring on the same day as of marking
were not included in the analyses to avoid underesti-
mating dispersal (Gall 1984). Each individual was
handled with all possible care, as marking and handling
individuals may affect their mobility. All individuals
received the same treatment during the marking pro-
cedure, on average 9 s. In addition to the MRR pro-
gramme all seminatural grasslands were surveyed for
butteries and burnet moths every 14 days in 2003
and 2004 (Nilsson & Franzn 2006). The number of
individuals, marked individuals and their marking
code was registered.

sins:xiii Nc

The effect of the study area size on estimated dispersal
distances was explored by a repeated subsampling
procedure using pre-dened smaller sections of the
study area. The subsampling procedure was performed
to sequentially reduce the area and randomly place the
subsampled area within the original study area. It was
a prerequisite that each subsampled area t completely
inside the original study area and that there were
recaptures inside each subsampled area. From each
subsampled area information about recaptures and
dispersal events that had occurred in the subsampled
area was extracted/intersected. All recaptures and
dispersal events with both patch of marking and patch
of recapture inside the subsampled area were selected
and average movement distances, mean and maximum
dispersal distances were calculated for the two burnet
species. Dispersal events were those found in the
subsampled area and based on our empirical ndings.
This procedure was repeated 100 times for each
subsampled area, species and year. The subsampling
started with a rectangular study area with the side lengths
10

8 km. By reducing the side length in 5% intervals,
and then examining which dispersal events that occurred
within the subsampled study area, the effect of the size of
the study area on dispersal distances could be explored.
Reducing the side length in percentage intervals does
not result in identical reduction of the study area size.
For example, subsampling using a 5% shorter side-
length result in a new study area covering 90% of the
original area, and a 90% reduction of side length result
in a new study area covering 1% of the original area.
The mean distance from the centre to the edge of
an average habitat patch was approximately 40 m.
Thus, we calculated the mean distance moved for all
recaptured individuals, assuming a movement distance
of 40 m for the individuals that were recaptured within
the same patch as where marked. A within patch
movement of 40 m corresponds well with other studies
on burnet moths (Bourn 1995; Kreusel 1999; Menndez,
Gutirrez & Thomas 2002). For the three individuals
that dispersed between more than two patches only
the longest distance was considered in the calculation
of mean dispersal distances. In the subsampling
procedure all dispersal events were included. The
subsampling of moved mean distance was based on
the number of recaptured individuals at each patch.
When the exact number of recaptured individuals
was unknown (because of the marking procedure) the
lowest possible number of individuals recaptured was
used. This was the case when several individuals from
the same patch and date were recaptured at several
later marking occasions. The minimum number of
recaptured individuals did not differ substantially
from the mean and maximum number of recaptured
individuals. For transferring individuals we used the
actual dispersal distance, counted as the shortest
distance between the two habitat patches.
We subsampled: (1) the mean movement distance of
all recaptured individuals for respective species in 2003
and 2004 (where within patch movement was set to
40 m); (2) mean distance travelled for individuals
transferring between different habitat patches (between
patch movements called dispersal, i.e. actual dispersal
distances) for respective species and year; (3) maximum
dispersal distance for respective species and year.
For

Z. lonicerae

there was not enough individuals
recaptured in 2004 to obtain a useful result. For the
repeated subsampling procedures

x:1i:n

65 were used.

Results

In 2003 a total of 4202

Z. viciae

individuals were
marked and 944 individuals were recaptured. Of these,
1227

Required area to
measure dispersal

2007 The Authors.
Journal compilation
2007 British
Ecological Society,

Journal of Animal
Ecology

,

76

,



12241230

80 individuals (85%), had dispersed to another habitat
patch. Mean transfer distance was 1078 m and the
maximum dispersal distance 5561 m. Mean movement
distance was 95 m. In 2004 a total of 2361 individual

Z. viciae

were marked. The number of individuals
recaptured was 232, and of these nine individuals
(39%), had dispersed to another habitat patch. Mean
transfer distance was 1805 m and the maximum dispersal
distance was 4062 m. Mean movement distance was
108 m. Approximately 20% of the total number of
individuals in the populations were estimated to have
been marked (unpublished data).
For

Z. lonicerae

554 individuals were marked of which
117 were recaptured in 2003. Of these, 10 individuals
(85%), had dispersed to another habitat patch. Mean
transfer distance was 540 m and the maximum dispersal
distance 3499 m. Mean movement distance was 65 m.
In 2004, 184 individual

Z. lonicerae

were marked of
which 25 individuals were recaptured. One individual
(40%), had dispersed to another habitat patch. Mean
transfer distance as well as maximum dispersal distance
was 494 m. Mean movement distance was 54 m.
Dispersal distances were not related to the number
of days between the day of marking and the day of
recapture (linear regression, B = 161,

P

= 0556,

r

2

= 0004). Mean distance moved, mean transfer
distance, and maximum dispersal distance increased
with size of subsampled study area (Fig. 3). For a study
area up to 10 km

2

, many dispersal events between
patches were undetected. There was an increase in
estimated dispersal distance as study area size increased
up to 5060 km

2

. In areas smaller than 30 km

2

, the
estimated dispersal measures were uncertain and low
(Fig. 3). For unbiased estimates of dispersal, areas of
50 km

2

or more are necessary for these species in the
study landscape.

Discussion

ni sirs:i s1ini rs

Most traditional dispersal studies are performed in
study areas that are too small, with species occurring
in low abundance resulting in dispersal patterns that
are often hard to interpret (Koenig

et al

. 1996; Wilson
& Thomas 2002). Among other organisms relatively
unbiased measurements of dispersal are available for
some birds (Newton, Davis & Davis 1989; Serrano

et al

. 2001), mammals (Roper, Ostler & Conradt 2003;
Haythornthwaite & Dickman 2006), amphibians (Smith
& Green 2006), reptiles (Rivera, Gardenal & Chiarav-
iglio 2006) and noctuid moths (William

et al

. 1993;
Nieminen 1996), where radio-tracking, genetic studies,
or large study areas reveal longer dispersal distances
than previously. Indeed, some species are very mobile
and require much larger study areas than our result
suggests. Some recent studies of birds have used very
large study areas, e.g. 10 000 km

2

(Serrano

et al

. 2001;
Hansson, Bensch & Hasselquist 2002). Although some
insects may only move a few hundred meters (Ingvar-
sson & Olsson 1997; Ranius 2006) and do not require
large study areas for estimates of dispersal distances.

vn:1 s1inx :r: si zr :r rqii rn:

Burnet moths are generally thought to be sedentary
and movements over 300 m are rarely found (Menndez

et al

. 2002; Young & Barbour 2004). However, we
found a strong relation between the size of the sub-
sampled study area and estimated dispersal distances.
We can put our data in perspective by comparing
with other dispersal studies of burnet moths. Mean
movement distances vary considerably among different
studies, but are positively correlated with the size of
the study areas (Table 1). The largest study area
demonstrated the longest dispersal distances (Kreusel
1999), a pattern that Schneider (2003) also showed for
butteries. Other studies of burnet moths with similar
study area sizes as in our subsamples, report similar
dispersal distance estimates, suggesting that our results
are not exceptional (Table 1).
There is a possible effect of the spatial conguration
of the habitat patches (i.e. connectivity and the degree
Fig. 3. Estimated movement and dispersal distances in relation to the study area size.
(ac) Mean distance moved (in metres) using all recaptures within subsampled study
areas for, respectively, size-class, species and year. All within habitat patch movements
was set to 40 m. (a) Zygaena viciae in 2003, (b) Z. viciae in 2004 and (c) Z. lonicerae in
2003. (df ) Mean transfer distance using all recaptures of individuals transferring
between patches within each subsampled study area for, respectively, size class, species
and year. (d) Z. viciae in 2003, (e) Z. viciae in 2004 and (f ) Z. lonicerae in 2003. (gi)
Maximum dispersal distance using the longest between patch dispersal event within
each subsampled study area for, respectively, size class, species and year. (g) Z. viciae in
2003, (h) Z. viciae in 2004 and (i) Z. lonicerae in 2003. Error bars show the 95%
condence interval of the mean.
1228

M. Franzn &
S. G. Nilsson

2007 The Authors.
Journal compilation
2007 British
Ecological Society,

Journal of Animal
Ecology

,

76

,



12241230

of fragmentation) on the dispersal pattern (Conradt

et al

. 2000; Baguette

et al

. 2003). Increased fragmenta-
tion increases the size of the study area required for
adequate dispersal measures as individuals may move
until they encounter suitable habitat or die (cf. Paradis

et al

. 1998). The required study area size that was
found is larger than study areas used in most previous
dispersal studies of insects, and it seems that a study
area of at least 50 km

2

is required for adequate dispersal
measures in some small terrestrial insects.
In favourable situations (e.g. suitable, warm and
sunny weather, large population sizes, and a large study
area) dispersal distances longer than normal could be
expected (Walls, Kenward & Holloway 2005). The
longest distance between occupied burnet habitat
patches in our study area was 9300 m. The longest
detected dispersal event was 5561 m and this distance
probably does not reect the absolute maximum
dispersal distance for these species. Our subsampling
procedure indicates, however, that the actual maximum
dispersal distance for these two species is close to the
distance found in this study (Fig. 3). However, the
distance of the longest dispersers will probably never be
captured in any traditional MRR study. The metrics
used to measure the longest dispersal events should
preferably be called the maximum dispersal event
measured.

:r rvrN1s :Nn :r sirci rs

To detect long distance dispersal events both a large
study area and large populations (i.e. large number of
recaptured individuals) are necessary. This is obvious
for

Z. viciae

in 2004, for which results from the sub-
sampling procedure is more variable, compared with
2003 (Fig. 3). If the two species studied had shown an
equivalent number of individuals, a low variation in
dispersal distance would have been expected. Most
species of conservation interest occur in small popula-
tions and small habitat patches, which constitute a
major problem for traditional dispersal studies. MRR
studies usually handle less than 1000 individuals
(Schtickzelle & Baguette 2003). For species with pop-
ulations limited to only a few localities MRR studies
should primary aim to estimate the local population
sizes and within patch movement behaviour.
How should we measure dispersal of rare species
when large areas and large sample sizes are required?
Indirect measures such as detailed mapping of the
distribution of a species over a longer period of time
has been suggested by Wilson & Thomas (2002), but
such studies are also to the outmost time consuming
and for many species the required information is not
available. Genetic methods are more frequently applied
to estimate dispersal and this eld might offer suitable
solutions of dispersal measures in the future (Nathan

et al

. 2003; Rivera

et al

. 2006). Radio telemetry can be
used on larger species (Hedin & Ranius 2002). How-
ever, direct traditional dispersal studies gain a lot of
information about local population structure that is
important. Our results emphasize that dispersal
patterns gained from small areas and small populations
should be interpreted with care.

Acknowledgements

Mikael Johannesson gave valuable help with the
subsampling procedure in

x:1i:n

. Charlotte Jonsson,
Sandra Rihm and Anneli hrstrm helped in the eld.
Thomas Gosden improved the English. Jens Roland
and one anonymous referee gave valuable comments
on an earlier version of the manuscript. Financial
support for this study was given by The Swedish
Research Council for Environment, Agricultural
Sciences and Spatial Planning (FORMAS).

References

Baguette, M., Mennechez, G., Petit, S. & Schtickzelle, N.
(2003) Effect of habitat fragmentation on dispersal in the
buttery

Proclossiana eunomia

.

Comptes Rendus Biologies

,

326

, 200209.
Table 1. Dispersal data obtained in markreleaserecapture studies of different burnet moths (Zygaena spp.)
Species
Mean dispersal
distance (m)
Maximum
dispersal
distance (m)
Study area
size (km
2
) Reference
Z. spp. 3300 13 Kreusel (1999)
Z. spp. 250400 1700 48 Grnwald (1988)
Z. spp. 1500 4 Naumann, Hille & Mller-Tappe (1987)
Z. spp. 100125 2000 < 1 Bhmer (1995)
Z. spp. < 200 980 < 1 Smolis & Gerken (1987)
Z. spp. 700 < 1 Bourn (1995)
Z. spp. 250 < 1 Lttmann (1987)
Z. lipendulae (L.) 33 533 < 1 Menndez et al. (2002)
Z. lipendulae 77 932 < 1 Schmidt-Loske (1993)
Z. viciae < 250 001 Young & Barbour (2004)
Z. viciae 95 5561 81 This study
Z. lonicerae 65 3499 81 This study
1229

Required area to
measure dispersal

2007 The Authors.
Journal compilation
2007 British
Ecological Society,

Journal of Animal
Ecology

,

76

,



12241230
Barrowclough, G.F. (1978) Sampling bias in dispersal studies
based on a nite area.

Bird Banding

,

49

, 331341.
Bhmer, T. (1995)

Untersuchungen zur abundanz, dispersion und
genetischen nachbarschaftsgre ausgewhlter populationen
des

Zygaena transalpina

superspezies komplexes (Insecta,
Lepidoptera, Zygaenidae)

. Diplomarbeit, Rheinische
Friedrich Wilhelms, Universitt, Bonn. [In German with
English summary.]
Bourn, N.A.D. (1995)

The ecology, conservation and population
genetics of three species of Zygaenid moths,

Zygaena
lonicerae

,

Zygaena purpuralis

and

Zygaena lipendulae

in
north west Scotland

. PhD Thesis, University of Aberdeen,
Aberdeen.
Conradt, L. Bodsworth, E.J. Roper, T.J. & Thomas, C.D.
(2000) Non-random dispersal in the buttery

Maniola
jurtina

: implications for metapopulation models.

Proceed-
ings of the Royal Society Biological Sciences Series B

,

267

,
15051510.
Englund, G. & Hambck, P.A. (2007) Scale dependence of
immigration rates: models, metrics and data.

Journal of
Animal Ecology

,

76

, 3036.
Gall, L.F. (1984) The effects of capturing and marking on
subsequent activity in Boloria acrocnema (Lepidoptera:
Nymphalidae), with a comparison of different numerical
models that estimate population size. Biological Conservation,
28, 139154.
Grnwald, V. (1988) Auswertung faunistisch kologischer
bestandsaufnahmen auf kalkhalbtrockenrasen bei Ottenbergen
(Kreis Hxter) unter besonderer bercksichtigung tagaktiver
schmetterlingsarten als beitrag zur pege und entwicklung-
splanung. Diplomarbeit, University of Paderborn, Hxter.
[In German with English summary.]
Hanski, I. (1994) A practical model of metapopulation
dynamics. Journal of Animal Ecology, 63, 151162.
Hanski, I. & Gaggiotti, O. (2004) Ecology, Genetics and
Evolution of Metapopulations. Elsevier Academic Press,
Amsterdam.
Hansson, B., Bensch, S. & Hasselquist, D. (2002) Predictors of
natal dispersal in great reed warblers: results from small and
large census areas. Journal of Avian Biology, 33, 311314.
Haythornthwaite, A.S. & Dickman, C.R. (2006) Long-
distance movements by a small carnivorous marsupial:
how Sminthopsis youngsoni (Marsupialia: Dasyuridae)
uses habitat in an Australian sandridge desert. Journal of
Zoology, 270, 543549.
Hedin, J. & Ranius, T. (2002) Using radio telemetry to study
dispersal of the beetle Osmoderma eremita, an inhabitant of
tree hollows. Computers and Electronics in Agriculture, 35,
171180.
Ingvarsson, P.K. & Olsson, K. (1997) Extinction-recolonization
dynamics in the mycophagous beetle Phalacrus substriatus.
Evolution, 51, 187196.
Koenig, W.D., van Vuren, D. & Hooge, P.N. (1996) Detectability,
philopatry, and the distribution of dispersal distances in
vertebrates. Trends in Ecology and Evolution, 11, 514517.
Kreusel, B. (1999) Dispersionsdynamik von Widderchen
zwischen verinselten kalkmagerrasen in Sueddeutschland
unter besonderer beruecksichtigung von naturschutza-
spekten. Natur und Landschaft, 74, 255265. [In German
with English summary.]
Lttmann, J. (1987) Verteilung von widderchenpopulationen
(Procris statices L., Zygaena lipendulae L., Zygaena
meliloti Esp.) in einem lebensraummosaik. Verhandlungen
der Gesellschaft fr kologie, 15, 359364.
Menndez, R., Gutirrez, D. & Thomas, C.D. (2002) Migration
and Allee effects in the six-spot burnet moth Zygaena
lipendulae. Ecological Entomology, 27, 317325.
Nathan, R. (2005) Long-distance dispersal research: building
a network of yellow brick roads. Diversity and Distributions,
11, 125130.
Nathan, R., Perry, G., Cronin, J.T., Strand, A.E. & Cain, M.L.
(2003) Methods for estimating long-distance dispersal.
Oikos, 103, 261273.
Naumann, C.M., Hille, A. & Mller-Tappe, S. (1987)
Rahmenbedingungen fr die Erhaltung und Entwicklung Von
Populationen Bedrohter Lepidopteren. Bielefeld University,
Bielefeld. [In German with English summary.]
Naumann, C.M., Tarmann, G.M. & Tremewan, W.G. (1999)
Western Palaearctic Zygaenidae. Apollo Books, Stenstrup.
Newton, I., Davis, P.E. & Davis, J.E. (1989) Age of rst breeding
dispersal and survival of red kites Milvus milvus in Wales
UK. Ibis, 131, 1621.
Nieminen, M. (1996) Migration of moth species in a network
of small islands. Oecologia, 108, 643651.
Nilsson, S.G. & Franzn, M. (2006) Biodiversity at Linnaeus
birthplace in the parish of Stenbrohult, southern Sweden.
5. Butteries and burnet moths. Entomologisk Tidskrift,
127, 3955. [In Swedish with English summary.]
Paradis, E., Baillie, S.R., Sutherland, W.J. & Gregory, R.D.
(1998) Patterns of natal and breeding dispersal in birds.
Journal of Animal Ecology, 67, 518536.
Ranius, T. (2006) Measuring the dispersal of saproxylic insects:
a key characteristic for their conservation. Population
Ecology, 48, 177188.
Rivera, P.C., Gardenal, C.N. & Chiaraviglio, M. (2006)
Sex-biased dispersal and high levels of gene ow among
local populations in the argentine boa constrictor, Boa
constrictor occidentalis. Austral Ecology, 31, 948955.
Roper, T.J., Ostler, J.R. & Conradt, L. (2003) The process of
dispersal in badgers Meles meles. Mammal Review, 33,
314318.
Schmidt-Loske, K. (1993) Spatial distribution of and habitat
use by Zygaena lipendulae (Linnaeus, 1758) (Lepidoptera,
Zygaenidae) in a suburban nature reserve. Proceedings of
the 5th International Symposium on the Biology of the
Zygaenidae (Insecta, Lepidoptera) (eds W.G. Tremewan,
W. Wipking & C.M. Naumann), pp. 165175. Koeltz
Scientic Books, Knigstein.
Schneider, C. (2003) The inuence of spatial scale on quantify-
ing insect dispersal: an analysis of buttery data. Ecological
Entomology, 28, 252256.
Schtickzelle, N. & Baguette, M. (2003) Behavioural responses
to habitat patch boundaries restrict dispersal and generate
emigration-patch area relationships in fragmented
landscapes. Journal of Animal Ecology, 72, 533545.
Serrano, D., Tella, J.L., Forero, M.G. & Donazar, J.A. (2001)
Factors affecting breeding dispersal in the facultatively
colonial lesser kestrel: individual experience vs. conspecic
cues. Journal of Animal Ecology, 70, 568578.
Shreeve, T.G. (1992) Monitoring buttery movements.
The Ecology of Butteries in Britain (ed. R.L.H. Dennis),
pp. 120138. Oxford University Press, Oxford.
Smith, M.A. & Green, D.M. (2006) Sex, isolation and delity:
unbiased long-distance dispersal in a terrestrial amphibian.
Ecography, 29, 649658.
Smolis, M. & Gerken, B. (1987) Zur frage der populationsgrsse
und der intrapopularen mobilitt von tagiegenden
schmetterlingen, untersucht am beispiel der Zygaenide-
narten (Lepidoptera: Zygaenidae) eines halbtrockenrasens.
Decheniana, 140, 102117. [In German with English
summary.]
Visnen, R. & Somerma, P. (1993) Suomen punatplper-
hoset (Zygaenidae of Finland). Baptria, 18, 149.
Walls, S.S., Kenward, R.E. & Holloway, G.J. (2005) Weather
to disperse? Evidence that climate conditions inuence
vertebrate dispersal. Journal of Animal Ecology, 74, 190
197.
William, B.S., Armon, J.K., Jimmy, R.R., William, H.H.,
Derrick, M.E., Michael, D.M., John, F.R., Michael, O.W.,
Michael, J.W. & John, L.G. (1993) Mechanism of southward
1230
M. Franzn &
S. G. Nilsson
2007 The Authors.
Journal compilation
2007 British
Ecological Society,
Journal of Animal
Ecology, 76,
12241230
migration of a noctuid moth Agrotis Ipsilon (Hufnagel): a
complete migrant. Ecology, 74, 23032314.
Wilson, R.J. & Thomas, C.D. (2002) Dispersal and the spatial
dynamics of buttery populations. Dispersal Ecology
(eds J.M. Bullock, R.E. Kenward & R.S. Hails), pp. 257
278. Blackwell Publishing, Oxford.
Young, M.R. & Barbour, D.A. (2004) Conserving the
New Forest burnet moth Zygaena viciae (Denis and
Schiffermueller) in Scotland; responses to grazing
reduction and consequent vegetation changes. Journal of
Insect Conservation, 8, 137148.
Received 20 February 2007; accepted 31 May 2007
Handling Editor: Stan Boutin