1 ALEKSANDRAS STULGINSKIS UNIVERSITY LITHUANIAN RESEARCH CENTRE FOR AGRICULTURE AND FORESTRY Rytis Zizas THE INFLUENCE OF FOREST STRUCTURE ON THE CAPERCAILLIE (TETRAO UROGALLUS L.) HABITATS AND DISTRIBUTION IN SOUTH- EASTERN BALTIC SEA REGION Summary of doctoral dissertation Agricultural sciences, Forestry (04 A) Akademija, 2015
40
Embed
ALEKSANDRAS STULGINSKIS UNIVERSITY LITHUANIAN …The dissertation is available at Martynas Mažvydas National Library of Lithuania and libraries of Aleksandras Stulginskis University
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
ALEKSANDRAS STULGINSKIS UNIVERSITY
LITHUANIAN RESEARCH CENTRE FOR AGRICULTURE
AND FORESTRY
Rytis Zizas
THE INFLUENCE OF FOREST STRUCTURE ON THE CAPERCAILLIE
(TETRAO UROGALLUS L.) HABITATS AND DISTRIBUTION IN SOUTH-
EASTERN BALTIC SEA REGION
Summary of doctoral dissertation
Agricultural sciences, Forestry (04 A)
Akademija, 2015
2
The doctoral dissertation was prepared at the Institute of Forestry, Lithuanian
Centre for Agriculture and Forestry in 2009-2014.
Scientific advisor:
Prof. dr. Gediminas Brazaitis (Aleksandras Stulginskis University, Agricultural
Sciences, Forestry 04 A);
The dissertation will be defended at Forestry Board of Lithuanian Research
Centre for Agriculture and Forestry and Aleksandras Stulginskis University:
Chairman:
Members:
Prof. dr. Vitas Marozas
Dr. Virgilijus Baliuckas
Opponents:
Dr. Rimgaudas Treinys
Dr. Marek Metslaid
The official defence of the dissertation will be held on April 1, 2015 at 10 a.m., in
4th building, meeting room 211, Aleksandras Stulginskis University.
Address: Universiteto st.11, Akademija, 53361 Kaunas district, Lithuania.
Summary of the doctoral dissertation was sent out on March 31, 2015
The dissertation is available at Martynas Mažvydas National Library of Lithuania
and libraries of Aleksandras Stulginskis University and Institute of Forestry,
Lithuanian Research Centre for Agriculture and Forestry
3
LIETUVOS AGRARINIŲ IR MIŠKŲ MOKSLŲ CENTRAS
Rytis Zizas
MIŠKO STRUKTŪROS ĮTAKA KURTINIŲ (Tetrao urogallus L.)
BUVEINIŲ PASIRINKIMUI IR PASISKIRSTYMUI BALTIJOS JŪROS
PIETRYČIŲ REGIONE
Daktaro disertacijos santrauka
Žemės ūkio mokslų sritis, miškotyros mokslo kryptis (04 A)
Akademija, 2015
4
Disertacija rengta 2009–2014 m. Lietuvos agrarinių ir miškų mokslo centro filiale
Miškų institute.
Mokslinis vadovas:
Prof. dr. Gediminas Brazaitis – Aleksandro Stulginskio universitetas, Žemės ūkio
mokslų sritis, Miškotyros kryptis (04 A);
Disertacija bus ginama Aleksandro Stulginskio universiteto Miškotyros mokslo
krypties taryboje:
Pirmininkas:
Prof. dr.
Nariai:
Prof. dr. Vitas Marozas
Dr. Virgilijus Baliuckas
Oponentai:
Dr. Rimgaudas Treinys
Dr. Marek Metslaid
Disertacija bus ginama viešame Miškotyros mokslo krypties tarybos posėdyje 2015
m. ... .. d. 10 val. Aleksandro Stulginskio universiteto IV rūmų posėdžių salėje
(211 kab.).
Adresas: Universiteto g. 8A, Akademijos mstl. 53361 Kauno r.
Disertacijos santrauka išsiuntinėta 20... m. ...mėn. ... d.
Disertaciją galima peržiūrėti Nacionalinėje Martyno Mažvydo, Aleksandro
Stulginskio universiteto ir Lietuvos agrarinių ir miškų mokslų centro filialo Miškų
instituto bibliotekose.
5
INTRODUCTION
The capercaillie densities have declined remarkably throughout its entire range,
especially in the past several decades (Wegge 1979; Miettinen, 2009). The species
is at the risk of extinction in the western, central and south-eastern Europe (Storch,
2007). For this reason, capercaillie is listed in Annex I of the EC Birds Directive
(Council Directive, 2009/147/EC), and the population trend appears to be
decreasing (IUCN, 2012; BirdLife International, 2015). Capercaillie has been listed
in the Red Data Book of Lithuania since 1989 (Rašomavičius, 2007) and attributed
to the rarity category II (critically endangered species). In accordance with Article
4 of the EC Birds Directive, Lithuania has established eight Special Protection
Areas (SPAs) for birds of this species.
Lithuania is located at the southwestern edge of a contiguous distribution range
of capercaillie. The species is fragmentary common in the southwestern part of
Europe. Therefore, it is necessary to ensure the protection of capercaillie habitats in
Lithuania avoiding the isolation of capercaillie populations. In addition, Lithuania
faces with stability problems of capercaillie leks (mating sites). Despite strong
fidelity and territoriality of capercaillie lek sites, leks tend to "move" or some areas
are abandoned by capercaillies and birds start using others leks. All known leks are
under protection in forests although birds have retreated and do not use them for a
long time. Changes in the forest structure could be one of the reasons why some
leks are eventually abandoned.
The abundance of birds depends on natural and anthropogenic factors, which
have certain influence at several spatial scale levels. Recent studies show that the
capercaillie, being a narrowly specialized and stand spatial structure demanding
species (Kurki and Lindén, 1995; Kurki et al., 2000; Graf, 2005), is sensitive to the
changes of macrohabitat at the landscape level. Various methods are used to
protect capercaillies in many countries as it is not known what particular features
of the habitat at the landscape level could create favourable conditions for
capercaillies. It is argued that capercaillies are one of the most thoroughly
researched species in western European countries (Storch, 2007); however, the
results may not always be adapted due to different geographical and ecological
conditions or forest management characteristics.
Biology and ecology of capercaillies were thoroughly studied in Lithuania by
Logminas more than fifty years ago (1962). Later, studies were mainly limited to
the identification of the location of capercaillie leks and monitoring of the
abundance of local populations in Special Protection Areas for birds and beyond.
Therefore, only a systematic and methodology based research can show the actual
condition of capercaillies (Kurtinio apsaugos planas, 2012). Until recently, there
were no detailed ecological studies, which would reveal the characteristics of
capercaillie habitats at both the level of stand structure (microhabitat) and
landscape (macrohabitat) level. The thesis mainly focuses on these two spatial
scale levels.
6
The aims of the thesis were to identify the features of lek selection by
capercaillies, forest structure in habitats used seasonally and the dynamics of home
ranges depending on the time of the year.
The objectives of the research:
1. To evaluate the influence of forest landscape structure variables on the
occupancy of capercaillie leks and to assess regional forest structure
differences.
2. To identify variables influencing capercaillies leks selection and to
establish the significant distance.
3. To investigate the home range size of capercaillies and the distance from
leks these birds live at different time of the year.
4. To analyse forest structure in capercaillie microhabitats used at different
time of the year.
Defended statements:
The occupancy of capercaillie leks are affected by forest structure.
Capercaillies select places for lekking in unstocked mature pine stands
near wetlands.
The dynamics of home range size of capercaillie males is related to the
time of the year and the period of life.
Capercaillies use various habitats depending on sex and the time of the
year.
Scientific novelty and practical importance
Selection of leks by capercaillies has been analysed for the first time in
Lithuania and the dependence of lek occupancy on the structure of forest landscape
has been ascertaine Studies on the use and movement of capercaillie habitats by
applying radio telemetry method for individual birds have been carried out for the
first time in Lithuania and Belarus.
The biological characteristics, which were identified during the studies,
substantially supplement the theoretical knowledge about this species in the
southwestern edge of continuous range. The obtained results can be used for
further studies on the interaction of these birds with the environment. The obtained
data on the microhabitats use at different time of the year can be used to
supplement forest felling regulations, which regulate forest management
capercaillie lek protection areas. The obtained data on the needs of capercaillies for
leks helped and will help in future to find as-yet-unknown leks and to efficiently
plan capercaillie population protection in Lithuania.
Volume and structure of the work
The dissertation is written in the Lithuanian language. It consists of:
introduction, literature review, material and methods, results and discussion,
conclusions, further needs for research and study trends, proposals for capercaillie
protection, reference list, list of publications, list of abbreviations and annexes. The
dissertation consists of 119 pages, including 15 tables, 57 figures, 193 references
and 3 annexes.
7
2. MATERIALS AND METHODS
2.1. Analysis of the influences of forest structure on the capercaillie leks
selection
Determination of leks and their occupancy level. The data on capercaillie
leks were obtained in 2008 conducting a questionnaire survey of employees of
state forest enterprises and regional and National parks. The locations, which were
indicated by the respondents, were adjusted to the data on the location of
capercaillie leks of State Forest Service (SFS). Seventy-one leks were chosen for
further investigation in total. Occupancy level and the location of lek centre were
determined as precisely as possible during the visits in March - April 2008-2012.
Lek occupancy level and lek centre were identified by seen/heard capercaillies,
and/or their activity signs. Leks were divided into ones of the high (≥2 males) and
low occupancy (attributed to individual male leks and non-used leks but still stored
in the SFS database). The new lek was distinguished, when it was >1000 m away
from the lek, including it in the SFS database. Whereas if the distance was ≤1000
m, the decision was made, that that was the same (continuous) lek (Rolstad and
Wegge, 1998a). GPS GARMIN device (the accuracy ±5 m) was used to record the
geographical coordinates of each lek centre. Forty-one leks (58 %) out of the total
71 capercaillie leks were attributed to the high occupancy leks, while 30 leks (42
%) were grouped as low occupancy leks. There were 21 leks of the high occupancy
(72 %) and 8 of low occupancy (28 %) in Dzūkija, while 22 leks of high
occupancy (52 %) and 20 (48 %) of low occupancy were distinguished in
Aukštaitija (Fig. 1).
8
Fig. 1. The distribution of high- (black spots) and low-occupancy (white spots)
capercaillie leks in south-eastern Lithuania
A database of geographical locations of 71 capercaillie leks was created. Three
buffer zones, keeping a distance of 2-5, 5-10 and 10-20 km, were established
around each point, indicating the lek. Random points were generated within these
zones using ArcGIS random point generator. Circular restricted areas of 0.25 km
(represent 19.6 ha), 0.5 km (78.5 ha), 1 km (314.2 ha) and 2 km (1256.6 ha) radii
were established around the centre of each lek and random points. The variables of
the environment describing forest structure at the stand and landscape levels,
including human disturbance, were identified in these territories as follows:
distribution of stands by tree species; pine forest age; stocking level of the stands;
and submitted them during the period from 2012 to 2013. The locations indicated
in the questionnaires were checked during the fieldwork according to the GPS
coordinates.
The data of 526 locations of capercaillies obtained in Lithuania (Dzūkija and
Aukštaitija) and Belarus (the Poozyorskaya forest) during the studies (from 2008 to
2013) were used in further studies. Two hundred forty (240) locations of
capercaillies were detected by radio telemetry method (50 locations in Lithuania);
259 capercaillies and/or locations with activity signs were found by detection (197
were found in Lithuania and 62 in Belarus); and 27 locations of capercaillies were
found using a questionnaire method (all these locations were found in Lithuania).
Data analysis. In all detected capercaillie microhabitats the geographic
coordinates were recorded using the GPS device and the following forest structure
parameters within 50 meters around were described: tree species composition, age,
the stocking level, ground layer, ecotones and the objects of infrastructure. Stand
species composition and site humidity were assessed visually. Since stand species
composition in capercaillie locations was very diverse at different time of the year,
stands were grouped according to the aforementioned indicators by distinguishing
10 categories of biotopes (Balčiauskas et al., 2008). The names of the biotope were
given according to the dominate tree species in the stand. The data was processed
using EXEL MS software.
Comparative analysis was used in order to further examine stand structure in
the microhabitats of capercaillies and to compare the habitats used by birds of
different sexes. The information stored in the Belorussian forest inventory
databases was not available; therefore, only the microhabitats of capercaillies
detected in Lithuania were used in the study. First of all, the database of
geographical positions was created for all (n=274) microhabitats. One random
point was established for each microhabitat using ArcGIS random point generator.
The points were located in the buffer zone leaving a distance of 2-5 km from the
centre of microhabitats. Circular area 50 m in radius (790 m2) was established
around each point marking the microhabitat and around the random point. The
following variables were identified in these areas: 1) dominant tree species in the
stand; 2) forest type; 3) site type; 4) pine stands age; 5) stocking level; 6) second
stand layer. The same databases, which were used to investigate lek surrounding,
were used in this investigation as well. The calculations of the areas of all the
aforementioned variables of stand structure were made using the statistical
software package SAS 9.3. The characteristics of stand structure in the
microhabitats of capercaillies (males and females separately) and the
characteristics of stand structure in random points were described and compared. In
order to evaluate the differences of microhabitats used by males and females, they
were compared with each other (and in individual regions as well). The differences
were tested using the Kolmogorov-Smirnov criterion (Zar, 2010), because data
samples did not meet normal distribution in most cases (checked by
UNIVARIATE procedure).
12
3. RESULTS AND DISCUSSION
3.1. Peculiarities of the capercaillies lek selection
In order to identify the peculiarities of the capercaillies lek selection, the
influence of individual variables of forest landscape structure on the occupancy of
leks and regional differences of forest structure were assessed first.
3.1.1. The influence of forest landscape structure variables to the
occupancy of capercaillie leks
Tree species composition. The proportion of pines in the stand in lek
surrounding and available landscapes was significantly different (Table 1). The
proportion of pine in available landscapes was significantly (p=0.004) lower if
compared with high occupancy leks, and did not differ significantly (p=0.17) if
compared with low occupancy leks. The proportion of pines in high occupancy
leks was significantly higher at the radii 0.25 km (p=0.029), 0.5 km (p=0.008) and
1 km (p=0.009) if compared with low occupancy leks.
The proportions of other four tree species (spruce, birch, black alder and aspen)
in lek surrounding and available landscapes were similar (Table 1). The proportion
of the aforementioned tree species in the stand did not differ significantly (p>0.05
in all the cases) in available landscapes if compared with both high and low
occupancy leks. Comparison of different occupancy leks revealed that the
proportion of all tree species was similar (p>0.05 in all the cases) regardless of the
analysed distance.
Table 1. Descriptive statistics of forests (the average ± SD, %) in capercaillie leks
surrounding (n=71) and available landscapes (n=71) at 1 km radius. Also the results of
statistical comparison of Kolmogorov-Smirnov (KS) test. Significant differences (p<0.05)
are in bold.
Variable Leks Available
landscapes KS test results
Tree species
composition
Pine 82.9 ± 11.2 70.9 ± 21.9 0.004
Spruce 6.5 ± 6.1 9.5 ± 6.5 0.054
Birch 8.6 ± 4.8 15.5 ± 9.4 0.084
Black alder 1.3 ± 2.0 3.2 ± 5.8 0.084
Aspen 0.1 ± 0.2 0.3 ± 0.7 0.26
Pine-stands
age
Clear-cuts 2.9 ± 2.8 4.1 ± 4.3 0.18
Young 9.4 ± 9.6 15.2 ± 9.6 0.12
Middle-aged 43.4 ± 15.2 33.4 ± 15.6 0.0012
Premature 23.1 ± 15.7 20.4 ± 14.3 0.185
Mature-overmature 21.2 ± 15.4 22.1 ± 16.0 0.99
Stands
stocking
level
Low (≤ 0.5) 4.8 ± 3.9 7.1 ± 5.1 0.084
Middle (0.6-0.7) 36.9 ± 16.4 36.9 ± 15.2 0.61
High (0.8-0.9) 55.2 ± 18.6 47.5 ± 16.6 0.0001
Very high (≥1.0) 3.1 ± 4.1 5.6 ± 7.5 0.034
13
Forest type
Vaccinio-myrtillosa 59.2 ± 26.7 55.8 ± 51.9 0.061
Cladoniosa 10.3 ± 15.4 13.9 ± 21.7 0.011
Ledo-sphagnosa 8.0 ± 13.4 7.1 ± 10.1 0.601
Carico-sphagnosa 6.3 ± 7.8 5.7 ± 6.3 0.026
Myrtillosa 5.9 ± 4.8 5.6 ± 7.9 0.151
Oxalidosa 5.2 ± 7.6 6.7 ± 9.6 0.004
Vacciniosa 4.4 ± 3.6 4.5 ± 4.1 0.090
Pine stands age. Middle-aged pine stands in lek surrounding covered a
significantly greater proportion of the area if compared with the available
landscapes. Comparison of the proportions of other maturity groups, as well as the
clear cuts (Table 1) revealed no significant differences The proportion of middle-
aged pine stands in available landscapes was significantly lower if compared with
high occupancy leks (p=0.004), and did not differ significantly (p=0.24) if
compared with low occupancy leks. The proportion of clear cuts, young stands,
pre-mature, mature-overmature pine stands in available landscapes did not
significantly differ if compared with both high and low occupancy leks (p>0.05 in
all the cases). Proportions of pine maturity groups as well as the clear cuts did not
differ significantly (p>0.05 in all the cases) on the leks with different occupancy
level regardless of the radius.
Stands stocking level. The proportions of low (≤0.5) and medium (0.6-0.7)
stocking level stands were almost the same in the lek surrounding and in available
landscapes. Meanwhile, the proportions of high (0.8-0.9) stocking level stands
were significantly larger in leks; while the proportions of very high (≥1.0) stocking
level stands were significantly smaller (Table 1). The proportion of low stocking
level in available landscapes did not differ significantly (p>0.05) if compared with
both high and low occupancy leks. The proportions of medium and high stocking
level in available landscapes were significantly smaller if compared with both high
and low occupancy leks (p<0.05 in all the cases). The proportions of very high
stocking level in available landscapes were significantly larger if compared with
both high and low occupancy leks (p<0.05). Comparison of different occupancy
showed that the proportions of all stocking level stands were similar (p>0.05 in all
the cases) regardless of the analysed distance.
Forest type. The proportions of Vaccinio-myrtillosa, Ledo-sphagnosa,
Myrtillosa and Vacciniosa stands in lek surrounding and in available landscapes
were similar (Table 1). Meanwhile, the proportions of Carico-sphagnosa stands in
leks were significantly higher, and the proportions of Cladoniosa and Oxalidosa
stands were significantly lower if compared with available landscapes (Table 1).
The proportions of Vm, M, V and Csp stands in available landscapes did not
differ significantly (p>0.05) if compared with both high and low occupancy leks.
The proportions of Lsp stands in available landscapes were significantly lower
(p=0.007) if compared with high occupancy leks and were significantly higher
(p=0.0003) if compared with low occupancy leks. The proportions of Ox stands in
available landscapes were significantly higher (p=0.009) if compared with high
occupancy leks and did not differ significantly (p=0.09) if compared with low
14
occupancy leks. Meanwhile, the proportions of Cl stands did not differ
significantly (p=0.37) if compared with high occupancy leks and were significantly
lower (p=0.007) if compared with low occupancy leks.
While comparing leks of different occupancy, the proportions of all forest type
stands within radii 0.25 km and 0.5 km did not differ significantly (p>0.05 in all
the cases). However, significantly larger proportions of Lsp stands were typical in
the leks of high occupancy in radii of 1 and 2 km (p<0.04 and p=0.002,
respectively) if compared with low occupancy leks. Meanwhile, the proportions of
Cl, Vm and Ox stands were larger significantly (p<0.05 in all the cases) in low
occupancy leks.
Forest land covered on average 87.9 ± 2.5% (SD) and 76.9 ± 16.5% (SD) of
the area in lek surrounding and in available landscapes, respectively (significant
difference, p<0.0001). Forest land proportions in high and low occupancy leks
surrounding were significantly larger (p<0.0001) if compared with available
landscapes. Comparison of the three radii up to 1 km from the centre revealed that
high and low occupancy leks did not differ from each other (p>0.05 in all the
cases). However, forest land proportion in high occupancy leks at 2 km in radius
was approximately 5% larger (significant difference, p=0.006).
Stands covered a similar proportion in lek surrounding and in available
landscapes (on average 95.6 ± 4.4 (SD) % and 94.8 ± 3.8 (SD) %, respectively)
(insignificant difference, p>0.05). No significant differences (p>0.05) were found
when comparing available landscapes with both high and low occupancy leks. No
significant differences were established when comparing leks of different
occupancy (p>0.05 in all the cases), regardless of the radius.
Core areas covered on average 79.8 ± 19.2 (SD) and 60.4% ± 22.7% (SD) of
the analysed area (p<0.0001) in lek surrounding and in available landscapes,
respectively. Core area proportions in high and in low occupancy leks were
significantly larger (p<0.0001) if compared with available landscapes. Core area
proportion in high and low occupancy leks did not differ significantly (p>0.05 in
all the cases), regardless of the radius.
Distance to the nearest forest edge. Leks (when measured from the centre)
were located about 350 m further from the forest edge (the average distance was
1470.3 ± 721.5 (SD) m) than available landscapes (1120 ± 301.5 (SD) m)
(significant difference, p<0.0001). High and low occupancy leks were significantly
(p<0.0001 in both cases) further from the forest edge than available landscapes.
Comparison of different occupancy leks revealed that high occupancy leks were
located about 180 m further from the forest edge (insignificant difference, p>0.05).
Open areas covered on average 2.3 ± 1.2 (SD) % and 3.8 ± 8.9 (SD) % in lek
surrounding and in available landscapes, respectively (insignificant difference,
p=0.54). Comparison of available landscapes and both high and low occupancy
leks revealed no significant differences. In addition, different occupancy leks did
not differ significantly from each other (p>0.05 in all the cases), regardless of the
radius.
15
Drained wetlands covered on average 4.1 ± 5.1 (SD) % and 6.2 ± 8.9 (SD) %
of research area in lek surrounding and in available landscapes, respectively
(insignificant difference, p=0.36). No significant differences (p>0.05) were found
when comparing available landscapes with both high and low occupancy leks. In
addition, different occupancy leks did not differ significantly from each other
(p>0.05 in all the cases), regardless of the radius.
Wetlands covered on average 15.6 ± 22.2 (SD) % and 14.1 ± 18.1 (SD) %, in
lek surrounding and in available landscapes, respectively (insignificant difference,
p=0.36). No significant differences were found when comparing available
landscapes with both high and low occupancy leks (p>0.05). Meanwhile, the
comparison of different occupancy leks revealed that the areas of wetlands were
significantly larger by 2.5-3 times in high occupancy leks (p<0.05 in all the cases)
at all analysed radii (Fig.2).
Fig. 2. The proportion of wetlands in the available landscapes (n=71) and low
(n=30) and high (n=41) occupancy capercaillie leks depending on radius from center
The distance to the nearest road with pavement from leks and available
landscapes (when measured from the centre) was about 815.7 ± 604.1 (SD) m and
590.2 ± 494.2 (SD) m, respectively (insignificant difference, p=0.08). High
occupancy leks kept significantly (p<0.01) longer distance, while low occupancy
leks kept insignificantly (p>0.78) longer distance to the road with pavement if
compared with available landscapes. High occupancy leks maintained about 260 m
(insignificant difference, p=0.53) longer distance to the road with pavement
(approx. 690 m).
The density of roads with pavement was on average 452.9 ± 430.6 (SD)
m/km² and 682.8 ± 603.8 (SD) m/km² (p=0.054) in lek surrounding and available
landscapes, respectively. The density of roads with pavement in available
landscapes was significantly (p<0.05) higher if compared with high occupancy
leks, while the differences were insignificant (p>0.05) when comparing with low
occupancy leks. No significant differences were found when comparing different
occupancy leks (p>0.05 in all the cases), regardless of the radius, although the
16
density of roads with pavement at 0.25 km and 0.5 km in radii in high occupancy
leks was about 1.5-2 times lower.
The distance to the nearest forest road from the capercaillie leks and
available landscapes was on average 480.7 ± 778.8 (SD) m and 590.3 ± 1132.6
(SD) m, respectively (insignificance difference, p=0.36). Both high and low
occupancy leks kept insignificantly longer (p>0.05) distance to the forest road than
available landscapes. High occupancy leks were about 560 m further from the
forest road, while low occupancy leks kept a distance of about 640 m to the forest
road (insignificant difference, p>0.71).
The density of forest roads was on average 1640.4 ± 1034.7 m/km² and
1504.6 ± 961.0 (SD) m/km² in lek surrounding and available landscapes,
respectively (insignificant difference, p>0.48). No significant differences (p>0.05)
were found when comparing available landscapes with both high and low
occupancy leks. Comparison of different occupancy leks showed that the density of
forest roads did not differ significantly in them (p>0.05 in all the cases), regardless
of the radius.
The distance to the nearest homestead from the capercaillie leks and
available landscapes was on average 1850.3 ± 1080.9 m (SD) and 1248.4 ± 989.4
m (SD) (significant difference, p<0.0001). High occupancy leks kept significantly
(p<0.0001) longer distance to the nearest homestead, while low occupancy leks
kept insignificantly (p=0.218) longer distance to the nearest homestead if
compared with available landscapes. Study revealed that high occupancy leks were
located much further from a homestead (significant difference, p=0.04) (approx.
2040 m) than low occupancy leks (approx. 1450 m).
3.1.2. Regional differences
Complex comparison (Table 2) showed that lek surrounding of different
regions differed in about one-third of all the analysed elements. The highest
number of differences between the regions was found in available landscapes,
located 2-5 km from the leks. The number of significant differences in the
variables of forest structure in available landscapes located 5-10 km from leks (if
compared with the differences in leks) decreased more than twice. No differences
were found in available landscapes located 10-20 km from leks. Mantel test results
revealed a declining correlation between the geographic distance and all the
analysed variables receding from lek centre. The obtained correlation results were
significant for leks (R=0.36) and for available landscapes located 2-5 km from leks
(R=0.22) (Fig. 3 A, B). The correlation was not significant for available landscapes
located 5-10 km and 10-20 km (R=0.09 and -0.06, respectively) from leks (Fig. 3
C, D).
Table 2. Results of comparison of forest landscape structure at 1 km in radius in
Dzūkija leks (n = 29), Aukštaitija leks (n = 41) and in available landscapes located at 2-5, 5-
10 and 10-20 km from the lek centre. The number of available landscapes is equal to the
number of leks. Kolmogorov-Smirnov (KS) test was used. The significance level: *–
17
0.05>p>0,01, ** – 0.001>p<0.01, ***– <0.001. Empty cells mean that the differences were
not significant.
Variable
Leks
Available
landscapes
2-5 km
Available
landscapes
5-10 km
Available
landscapes
10-20 km
Forest land
Core area
Stands
Wetlands
Drained wetlands * *
Open areas
Roads with pavement ** **
Forest roads ** ***
Distance to:
Forest edge
Road with pavement *
Forest road * **
Homestead * *** *
Pine-dominated stands age:
Clear-cuts
Young ** **
Middle-aged ***
Premature * *
Mature-overmature *
Proportion of tree species
Pine * *
Spruce *** ** *
Birch * *
Black alder
Aspen
Proportion of stocking level
≤0.5 *
0.6-0.7
0.8-0.9
≥1.0
Proportion of forest type
Vaccinio-myrtillosa ** *
Cladoniosa * *
Oxalidosa ** ***
Vacciniosa *
Myrtillosa
Ledo-sphagnosa
Carico-sphagnosa
Regions mostly differed in the following variables: density of the roads with
pavement, density of forest roads, proportion of drained wetlands, distance to a
18
homestead and distance to forest road, proportion of middle-aged pine, spruce
stands, Vm and Ox forest type sites. The average canonical correlation coefficient
of all these variables was R=0.40.
A B
C D Fig. 3. Mantels test’s analysis of forest structure variables at 1 km radius in: leks
surrounding (A); available landscapes at 2-5 km distance from leks (B); available landscapes
at 5-10 km distance from leks (C); available landscapes at 5-10 km distance from leks (D)
3.1.3. The influence of forest structure on the capercaillie lek selection
The comparison of leks at 0.25, 0.5, 1 and 2 km in radius revealed that lek
environment differed up to 0.5 and from 0.5 km in radius most. Meanwhile, the
surrounding remained almost the same in the restricted areas at 0.25 km from 0.5
km and 1 from 2 km in radius. The analysis of the capercaillies preferred leks in
different regions within a distance of 0.5 km from the centre in high and low
occupancy leks revealed the influence of variables of the same forest landscape
structure. In addition, the influence of the same variables was also typical on the
selection of leks in different regions (Dzūkija and Aukštaitija). These variables
included: low (0.5), medium (0.6-0.7) and high (0.8-0.9) stocking level stands, pine
stands, middle-aged pine stands, mature-overmature pine stands, wetlands, stands
of Vm and Lsp forest type and core area (Fig. 4).
19
A B
C D Fig. 4. Forest landscape variables, influencing the capercaillie leks selection at 500 m radius
in Dzūkija (high-A; low-B occupancy leks) and Aukštaitija (high-C; low-D occupancy leks)
Meanings: stands (me); forest land (mi); wetlands (pe); drained wetlands (nu); open areas (ai); core area (mv); roads with pavement (kd); forest roads (mk); distance to the forest edge (mp); distance
to the road with pavement (ad); distance to the forest road (am); distance to the homestead (as); pine
stands (P); clear cuts (ki); young pine stands (P_jaun); middle- aged pine stands (P_pus), premature pine stands (P_prib); mature-overmature pine stands (P_br); spruce stands (E), young spruce stands
(E_jaun); middle-aged spruce stands (E_pus), old spruce stands (E_sen); birch stands (B); young birch
stands (B_jaun); middle- aged birch stands (B_pus), old birch stands (B_sen); cl, ox, lsp, csp, v, vm, m – stands of different forest types; _05, 0.6-0.7, 08-0.9, _10 – stands of different stocking level
Discussion. The influence of forest landscape structure on the occupancy of
leks of capercaillies. The lek surrounding differed from available landscapes in
about one third of all the analysed variables of forest landscape structure. The
obtained results confirm one of the defended statements, because more
(statistically significant) differences were found while comparing available
landscapes with high occupancy leks than comparing with low occupancy leks of
capercaillies. This shows the influence of specific variables to the occupancy level
20
of leks. The surrounding of low and high occupancy leks located at different
distance from the centre differed in the following variables: stand species
composition, forest type, areas of forest land and wetlands, and distance to forest
edge, to a homestead, to the roads with pavement and their density. Different
occupancy leks did not differ in the following variables: proportion of the clear
cuts and pine stand age, stocking level of the stands, proportion of the stands, core
areas, open areas, drained wetlands, density of the forest roads and distance to
them.
The proportion of forest land and forest depth in leks was significantly higher if
compared with the landscape around random points. This fact showed that leks
were spread throughout relatively less fragmented forest areas. This is in full
agreement with the opinion of other researchers who claim that leks are located in
continuous forest tracts (Rolstad and Wegge, 1987; Helle et al., 1994). Smaller
forest land proportion was found in low occupancy leks at 2 km in radius, while the
distance from low occupancy leks to forest edge was shorter. Stands proportion, in
contrast to the proportion of forest land, was smaller in high occupancy leks if
compared with low occupancy leks. This shows that a greater forest fragmentation
was typical to low occupancy leks than to high occupancy leks. Forest
fragmentation is considered to be one of the main factors negatively affecting the
abundance of capercaillie populations worldwide (Andrén, 1994). However, in
order to clarify the obtained results, research should be continued further.
The proportions of wetlands and stands of Ledo-sphagnosa forest type had a
tendency to be larger in high occupancy leks than in low occupancy leks at all
analysed radii. On the contrary, proportions of Vaccinio-myrtillosa, Myrtillosa,
Vacciniosa, Oxalidosa and other forest type stands were larger in low occupancy
leks. That shows a positive influence of wetlands on the occupancy of leks. Results
revealed that low occupancy leks were located closer to homesteads, roads with
pavement and the density of roads with pavement was higher there. This suggests
that anthropogenic disturbance sources are one of the most significant factors to
determine lek occupancy level. The obtained results are in full agreement with the
results of other authors, who indicate that the anthropogenic factors affect the
capercaillies’ choise of habitats (Leclercq, 1985; Ménoni and Bougerol, 1993;
Sachot et al., 2003).
Lek selection. Regardless of forest structure differences of two regions in the
surrounding of capercaillie leks and in the area up to 5 km in radius from leks, lek
selection was influenced by the same variables, which are as follows: stands of low
(≤0.5), medium (0.6-0.7) and high stocking level (0.8-0.9), pine stands proportion,
middle-aged, mature-overmature pine stands, wetlands, Vaccinio-myrtillosa and
Ledo-sphagnosa forest types and core area. The influence of the aforementioned
variables on lek selection was obvious at 0.5 km in radius from the centre of leks.
This shows lek uniqueness in the central part, which have the so-called male
display grounds. The influence of forest structure on lek selection was not revealed
within the distance of 1 and 2 km from the centre of leks, because these areas were
under the boundaries of daily territories of males. The surrounding of these areas
21
was less distinct if compared with display grounds. In addition, the same variables
were typical to both high and low occupancy leks. This tendency shows that the
differences of lek surrounding (or, in other words, habitat quality) can explain the
occupancy of leks.
Capercaillies prefer pine stands throughout their range (Logminas, 1962;
Seiskari, 1982; Черкас, 2008). The results of this analysis revealed that lek
selection was influenced by middle-aged and mature-overmature pine stands (when
considering stand age). The prior studies emphasized the importance of old forests
(e.g. Hjorth, 1970; Wegge and Rolstad 1986; Picozzi et al., 1992; Kurki et al.,
2000). There is no uniform opinion about the influence of pine stands of different
maturity groups on the lek selection of leks. It is said that mature stands, which
grow in the intensive management forests, lose some characteristics typical to
natural forests (Miettinen et al., 2009). In order to adapt to the changing conditions
(Dzieciolowski and Matuszhewski, 1980), capercaillies are forced to choose
younger forests (Mykrä et al., 2000; Miettinen et al., 2005; Sirkiä et al., 2011). The
results of this study as well as the results obtained by Sirkiä et al. (2011) in Finland
(leks were analysed at 3000 m in radius) show that lek occupancy did not depend
on the share of old stands. According to the data of the studies carried out by
Miettinen (2009), even the areas of young pine stands in managed forests in
Finland positively correlated with the males’ number at the lek.
It is likely that stands of high (0.8-0.9) stocking level had a negative influence,
because dense stands reduce the chances of capercaillies to escape from predators
(Rolstad and Wegge, 1990). In this paper, the stands of low (≤0.5) and high (0.6-
0.7) stocking level were identified as factors influencing lek selection. Other
authors also mentioned the preference of capercaillies in the thin stands. I. Storch
(1993) argued that approx. 2/3 of middle-aged stands in central Europe are too
dense for capercaillies, especially multi-storey stands (stands with more than one
storey). It is said that capercaillies use open areas for mating where visibility is 20-
50 m (Valkeajärvi and Ijäs, 1986; Miettinen, 2009). When collecting data, the
mating games of capercaillies were often detected on the forest roads, compartment
and electrical lines. These findings agree with the findings of other authors
(Dzieciolowski and Matuszewski, 1980).
The results of this study indicated that wetland areas (Ledo-sphagnosa forest
type), influenced lek selection and lek occupancy. Meanwhile, Vaccinio-myrtillosa
forest type stands were very important for after mating period.
22
3.2. The home range size of capercaillies
Eight radio-tagged capercaillies (7 males and 1 female) were tracked during the
study period. Two hundred eighty (280) locations of these birds were registered in
total. The size of the total annual home ranges used by individual birds and home
ranges used at different time of the year were defined based on the obtained data
(Table 3).
Spring. Birds were captured and radio-tagged at the end of their mating period
(the second half of April). The home range used by the male M303 amounted to
32.3 ha from April 23 to May 10, 2010. The bird moved 1.1 km the farthest from
the place of its capture. The area used by this bird in next year’s spring, from 1
April until mid-May, amounted to 51.2 ha, while 1.3 km was the maximum
distance from the lek (the place of capture). During the same period, the area used
by another male (M02) was slightly larger (63.8 ha) and the maximum distance
from the lek was 1.5 km (Table 3).
Table 3. The home range (HR) and movement distances of radio-tracked capercaillies (n=8)