Berkut. Vol. 8.
Is. 2. 1999. P. 150-154.
The impact of
landscape structure
on occurrence of
White Stork’s nests
C. Latus, K.
Kujawa
Влияние структуры ландшафта на встречаемость гнезд белого
аиста. - К. Латус, К. Куява. - Беркут. 8 (2). 1999. -
Целью работы было определение показателей структуры ландшафта, влияющие
на встречаемость гнезд белого аиста. Исследования проведены на участке в 660
км2 в Восточной Германии (долина р. Одер), где плотность гнездования
достигает 4 пар на 100 км2. Предпочтение местообитаний белым аистом
определялось путем исследования структуры ландшафта в радиусе 1 км от гнезда. В
населенных пунктах было случайным образом выбрано 56 “гнездовых участков” и 34
“контрольных участка” (без гнезд). Структура ландшафта (соотношение
местообитаний, индекс разнообразия Шеннона H', плотность границ и т. д.)
измерялись по аэрофотоснимкам (1:10000). “Гнездовые участки” характеризовались
более высокой долей травянистых биотопов и водоемов и большим значением H':
1,37 по сравнению с 0,95 на “контрольных участках” (P < 0,01). Аисты явно
предпочитали более разнообразные местообитания – 80 % “гнездовых участков”
имели высокую степень разнообразия (0,81-1,70), в то время как 80 %
“контрольных участков” в намного более низком классе разнообразия (0,21-0,80).
Травянистые биотопы также имели значительное влияние. Аисты предпочитали для
гнездования прилегающие и умеренно фрагментированные участки.
Abstract. The
objective of the study was to determine the landscape structure indices
influencing the occurrence of White
Stork’s (Ciconia ciconia) nests. The
analyses were performed for the MOL-district (660 km2) in East Germany
(Oder valley) where the density of breeding pairs as high as to 4 per 100 km2.
White Stork habitat preferences were determined by investigating landscape
structure within a 1 km radius of nesting locations, ‘breeding sites’ (N = 56),
within and for ‘control sites’ (N =
34), randomly selected points in the
villages without stork’s nests. The landscape structure (habitat proportions,
Shannon’s diversity index H’, density of edges, etc.) was measured by
evaluating aerial photographs (1:10000). Breeding sites were characterised by a
higher proportion of grasslands and inshore waters as well as by a
significantly higher (P < 0,01) value of H’ 1,37 compared to 0,95 for
‘control sites’. Storks clearly preferred more differentiated landscapes – 80 %
of ‘breeding sites’ were localised in very high diversity classes (0,81-1,70),
while 80 % of ‘control sites’ were in much lower diversity classes (0,21-0,80).
Also grasslands were found to have a strong influence. Storks preferred
contiguous and moderately fragmented grasslands as nesting habitats.
Key words: White Stork, East
Germany, habitat preference, landscape structure.
Address: Claudia
Latus, ZALF, Institute of land-use
systems and landscape ecology, Eberswalderstr. 84, 15374 Müncheberg,
Germany. e-mail: clatus@zalf.de.
Krzysztof Kujawa, Research Center for Agricultural and
Forest Environment, Polish Academy of Sciences, Field Station, Szkolna 4,
64-003 Turew, Poland. e-mail: ortolan@priv7.onet.pl.
Introduction
Intensive land-use management affects the structural arrangement of agricultural
landscapes and influences the occurrence of species leading to change or
serious loss of biodiversity (Mc Laughlin, Mineau 1995). Additional
information is needed concerning abiotic (physical and chemical) landscape
features and connections between landscape and the biota to help forecast ecological threats and
support efforts to protect biodiversity. The White Stork (Ciconia ciconia) was selected for an investigation of the
integration of a species into a landscape due to its large area-habitat
preferences and ability to adapt to farmland. The objective of our study was to
determine the landscape parameters strongly associated with the occurrence of White Stork’s nests and,
therefore those which should be considered as essential for effective
protection of the species.
Study
area
The study was carried out on a 660 km2 area in the district MOL (Brandenburg),
located in the Oder river valley. In the 1990s, the density of White Stork
reached as high as 4 pairs/100 km2
(Mitteilungsblatt..., 1995). The district MOL has a typical land use
composition for agricultural regions: agricultural use (62 % ) (92 % crop
fields and 7 % grasslands), woodlands (24 %), inshore waters (3 %), settlements
(9 %) and other land uses (2 %) (LDS Brandenburg, 1995). Fifty-one percent of
the crop fields is used for grain. The land is intensively cultivated: the
annual applications of NPK reach 200 kg per ha, the wheat yield is 45-55 dt/ha
and pesticides are sprayed from 2-6
times/year (Amt für Landwirtschaft, 1995). The district MOL has 82
inhabitants/km2. The density of human settlement is 6/100 km2.
The annual mean precipitation is approx. 500 mm; the mean temperature is +18 °C in July and –1 °C in January (LDS Brandenburg, 1995).
Methods
The landscape structure in the proximity of
nests occupied by breeding pairs within last five years (“breeding sites”: BS,
N = 56) was compared to the landscape
structure of areas without White Stork’s nests (“control sites”: CS, N = 34).
Because nearly all of the nests are
localised in the settlements (Ranner, Tiefenbach 1994), CS were taken as
randomly selected points in settlements without White Stork’s nests, close to the border of the settlements. The
landscape structure was investigated within a 1 km radius around BS and CS, based on a finding by Creutz (1988) who
noticed that during the earliest stage of fledgling development their parents
need view-contact with the nests. The landscape structure was measured with the
aid of black and white aerial photographs at a 1:10 000 sale. The measurements
were focused on the patches and boundaries between patches. A “patch” was
defined as a homogeneous area of land (according to photograph sensitivity) and
classified according to six land-use classes: woodlands, crop fields,
grasslands, inshore waters (seas, ponds etc.), bushes (cemeteries, gardens
etc.), settlements.
The
chosen variables can be split into three groups:
1.
Variables based on the main land-use classes
proportions: a) the percentages of crop fields, woodlands, grasslands,
inshore waters, bushes, settlements, b) landscape diversity index defined as
Shannon’s diversity index , in which
is the proportion of i-th habitat and
c) the openness of a landscape. The openness of landscape is a good indicator
for birds to determine a free view of the nest surroundings to recognise
possible feeding areas as well as dangers. We defined an open land as covered
by crop fields and/or grasslands (Andries, 1984), thus the openness of
landscape is the percentage of open land.
2.
Variables based on the linear structures: the densities of different types of
boundaries (between crops and grasslands, woodlands and grasslands, etc.) as
well as their total.
3. Variables based on the distances: a) the
distance from BS (CS) to the nearest patch determined separately for all
land-use classes, b) the distance between the same type of land-use patches.
Results
Impact
of land-use types composition
The
land-use of the area at BS and CS differed markedly. BS’s area was characterised
by a smaller percentage of crop fields (52 % and 71 %, respectively), and much
higher percentage of grasslands (11 % and 4 %) as well as inshore waters (6 %
and 1 %). According to Chi-square test the differences are statistically
significant at P<0,001 (Fig. 1). The difference in landscape diversity
between BS and CS (1,37 and 0,95,
respectively) is statistically significant at P<0,01 (Poole, 1979). In
addition, the difference in the frequency distribution between BS and CS in
relation to the H’ index is
statistically significant (cc2 test, P<0,001). Eighty percent
of BS were localised in high diversity classes of 0,8-1,7 while 80 % of CS
were found in lower diversity
classes of 0,2-0,8.

Fig. 1. Land-use of the area around breeding
sites (BS) and control sites (CS).
Рис. 1. Использование земель вокруг гнездовых (BS) и контрольных (CS)
участков.
Special consideration was given to some types of
land-uses important for White Storks for nesting and feeding purposes. Crop
fields and grasslands may be regarded as potential feeding areas but woodlands
as a possible indicator for disturbance. The frequency distributions of BS and
CS varied (cc2 test, P<0,001) in relation to crop field coverage
(Fig. 2a) as well as to grassland percentage (Fig. 2b). Sixty percent of BS
were localised in places characterised by crop fields covering areas in a range
of 20-60 %, while up to 80 % of CS were surrounded by areas covered by
60-100 % of crop fields. Seventy-five
percent of BS were found in the places where grasslands cover from 5 % to more than 30 % of the land
surface, while about 85 % of CS were found in areas with a very low percentage
(0-10 %) of grasslands. These results correspond with the grassland/open
land rate analyses which showed a significantly higher value of the rate (U-test, P<0,001) for BS. Regarding
woodlands, 80 % of BS were found in the areas with 0-20 % woodland
coverage, while 60 % of CS were
surrounded by 10 % to more than 30 % woodland coverage (Fig. 2c). The
differences are statistically significant (cc2 test, P < 0,001).

a)

b)

c)
Fig. 2. Frequency distribution of breeding
sites and control sites.
Рис. 2. Частотное распределение гнездовых и контрольных участков:
a) – в зависимости от площади полей;
b) – в зависимости от площади лугов;
c) – в
зависимости от площади лесов.
Impact
of boundaries
BS were more frequently localised in the range
of 12,1-24 km of all boundaries per circle while CS were found to be markedly
most frequent in the range of 12,1-16
km per circle (Fig. 3). The difference reported here is statistically
significant (cc2 test, P < 0,01). Special attention was focused
on the fragmentation of grasslands the main feeding areas for White Stork.
Nests were found more frequently in the areas characterised by significantly
higher density of grassland boundaries (cc2 test, P < 0,001)
(Fig. 3).


Fig. 3.
Frequency distribution of BS (thick line) and CS (thin line) in relation to
density of boundaries (in km/circle).
Рис. 3. Частотное распределение гнездовых (толстая линия) и контрольных
(тонкая линия) участков в зависимости от плотности границ (в км на круг).
Discussion
Most studies on the distribution of White Stork’s
nest were based on singular habitat analyses, for example on the influence of
grassland percentage on the
density of breeding pairs (Thomsen, 1995, Schneider, 1988). The results
reported here show much more complicated
relationships between the species and preferred
habitat and support previous findings that investigations of the species
must be carried out in landscape contexts as other authors have already noticed
(Flather, Sauer 1996).
The main factors responsible for survival of a
given species are the possibility for nest building and sufficient food supply
for its brood. The importance of food supply for White Stork in his breeding
region was showed e. g. by Pinowska et al. (1989) and the influence of food
supply on breeding success has been analysed by Löhmer et al. (1980),
Profus (1986), Struwe and Thomsen (1991). Although results are based on the analysis
of landscape structure characteristics, clear preferences for special landscape
features outside of the one km radius can be easily recognised. The landscape
structure around BS and CS differed depending on the percentages of crop
fields, grasslands and woodlands as well as on the diversity of landscape.
Indeed variables based on distances are of less importance within 1 km around
nest, but variables based on linear structures are significant different and provide more comprehensive information
about the degree of landscape fragmentation determined by the density of
boundaries. These preferences are surely linked with the feeding biology of
White Stork. The species is a generalist in food intake. As Lakeberg (1995)
described, this species changes its food during the breeding season from
earthworms and amphibians in the first
phase to small mammals preferred
later during breeding older fledglings, probably because of a higher energy
content (Profus, 1986). But decisive role for White Stork is played not by the
potential food supply by itself, but food availability (Alonso et al.,1991).
The species as a stepper bird with quick sight reacts on small moving animals
(Creutz, 1988). White Stork looks for mice and other animals in open land, especially
with low vegetation, harvested crop fields and grasslands, which can be found
much more easily (and on the average closer to the nest) in diversified
landscape compared to unified one. So, it seems that diversified, mosaic-like
landscapes consisting of many patches of different land-use types (incl. small
woods, small water bodies etc.) but with some minimum area of grasslands (10-20
%) as main feeding area and moderate density of all boundaries as well as
boundaries of grassland patches are an optimal breeding habitat for White Stork
in intensively used farmland in Europe.
Acknowledgement: The investigation was executed in co-operation between ZALF, Institute
of Land-use Systems and Landscape Ecology in Germany and RCAFE PAS in Poland
supported by BML (Project No. 84). Many thanks for the data on White Stork to N. Wenzel, W. Trebesch and W. Schmidt
from NABU, Germany as well as to M. Röhl for technical help.
references
Alonso J.C., Alonso J.A., Carrascal L.M. (1991): Habitat selection by
foraging White storks Ciconia ciconia
during breeding season. - Can. J. Zool. 69: 1957-1962.
Amt für Landwirtschaft (1995): Übersichtsinformation zum
strukturellen und wirtschaftlichen Entwicklungsstand der Landwirtschaft im
Landkreis Märkisch Oderland (11.10.1995 Amt für Landwirtschaft in
Neuhardenberg eds.).
Andries A.M., Gulinck H., Herremans M. (1994): Spatial modelling of barn
owl Tyto alba habitat using landscape
characteristics derived from SPOT data. - Ecography. 17 (3): 278-287.
Creutz G. (1988): Der Weißstorch. Die Neue Brehm-Bücherei.
375. Wittenberg Lutherstadt: A. Ziemsen Verlag. 1-236.
Flather C.H., Sauer J.R. (1996): Using landscape ecology to test
hypotheses about large-scale abundance pattern in migratory birds. - Ecology.
77: 28-35.
Lakeberg H. (1995): Zur Nahrungsökologie des Weißstorchs Ciconia ciconia in Oberschwaben
(S-Deutschland): Raum-Zeit-Nutzungsmuster, Nestlingsentwicklung und
Territorialverhalten. - Ökologie der Vögel. 17 (Sonderheft): 1-87.
LDS Brandenburg (1995): Statistische Jahrbücher des Landes
Brandenburg 1994-1997.
Löhmer R., Jaster P., Reck F.G. (1980): Untersuchungen zur
Ernährung und Nahrungsraumgröße des Weißstorches. -
Beitr. Naturkunde Niedersachsens. 33: 117-129.
Mc Laughlin A., Mineau P. (1995): The impact of agricultural
practices on biodiversity Agr. - Eco. Env. 55: 201-212.
Mitteilungsblatt der BAG Weißstorchschutz (1995): NABU. 87: 1-16.
Pinowska B., Pinowski J. (1989): Feeding ecology and diet of the White
Stork Ciconia ciconia in Poland. -
Weißstorch - White Stork. Proc. I Intern. Stork Conserv. Symp.
Schriftenreihe des DDA. 10: 381-396.
Poole R.W. (1979): An Introduction to Quantitative Ecology. New York:
Mc Graw-Hill. 1-533.
Profus P. (1986): Zur
Brutbiologie und Bioenergetik des Weißstorchs in Polen. - Beih.
Veröff. Naturschutz Landschaftspflege Bad.-Württ. 43: 205-220.
Ranner A., Tiefenbach M. (1994): Der Weißstorch: Bestandesentwicklung,
Gefährdungsursachen und Maßnahmenvorschläge. - Reports
Umweltbundesamt. Wien UBA-94-095.
Schneider M. (1988): Periodisch überschwemmmtes Dauergrünland
ermöglicht optimalen Bruterfolg des Weißstorches in der Save-Stromaue (Kroatien/Jugoslawien).
- Vogelwarte. 34: 164-173.
Struwe B., Thomsen K.M. (1991): Untersuchungen zur Nahrungsökologie
des Weißstorchs in Bergenhusen 1989. - Corax. 14: 210-238.
Thomsen K.M. (1995): Auswirkungen moderner Landbewirtschaftung auf die
Nahrungsökologie des Weißstorchs.- Proc. Intern. Symp. on the White
Stork (Western Population). Basel 1994. 121-134.